paper_id
string
title
string
paper_url
string
pdf_url
string
authors
list
abstract
large_string
track
string
primary_area
string
doi
string
volume
string
issue
string
pages
string
abstract_source
string
arxiv_id
string
arxiv_id_source
string
10.1609/aaai.v35i2.16264
What to Select: Pursuing Consistent Motion Segmentation from Multiple Geometric Models
https://ojs.aaai.org/index.php/AAAI/article/view/16264
https://ojs.aaai.org/index.php/AAAI/article/download/16264/16071
[ "Yangbangyan Jiang", "Qianqian Xu", "Ke Ma", "Zhiyong Yang", "Xiaochun Cao", "Qingming Huang" ]
Motion segmentation aims at separating motions of different moving objects in a video sequence. Facing the complicated real-world scenes, recent studies reveal that combining multiple geometric models would be a more effective way than just employing a single one. This motivates a new wave of model-fusion based motion ...
main
Computer Vision
10.1609/aaai.v35i2.16264
35
2
1708-1716
official
null
null
10.1609/aaai.v35i2.16265
Asynchronous Teacher Guided Bit-wise Hard Mining for Online Hashing
https://ojs.aaai.org/index.php/AAAI/article/view/16265
https://ojs.aaai.org/index.php/AAAI/article/download/16265/16072
[ "Sheng Jin", "Qin Zhou", "Hongxun Yao", "Yao Liu", "Xian-Sheng Hua" ]
Online hashing for streaming data has attracted increasing attention recently. However, most existing algorithms focus on batch inputs and instance-balanced optimization, which is limited in the single datum input case and does not match the dynamic training in online hashing. Furthermore, constantly updating the onlin...
main
Computer Vision
10.1609/aaai.v35i2.16265
35
2
1717-1724
official
null
null
10.1609/aaai.v35i2.16266
Deep Low-Contrast Image Enhancement using Structure Tensor Representation
https://ojs.aaai.org/index.php/AAAI/article/view/16266
https://ojs.aaai.org/index.php/AAAI/article/download/16266/16073
[ "Hyungjoo Jung", "Hyunsung Jang", "Namkoo Ha", "Kwanghoon Sohn" ]
We present a new deep learning framework for low-contrast image enhancement, which trains the network using the multi-exposure sequences rather than explicit ground-truth images. The purpose of our method is to enhance a low-contrast image so as to contain abundant details in various exposure levels. To realize this, w...
main
Computer Vision
10.1609/aaai.v35i2.16266
35
2
1725-1733
official
null
null
10.1609/aaai.v35i2.16267
Spectral Distribution Aware Image Generation
https://ojs.aaai.org/index.php/AAAI/article/view/16267
https://ojs.aaai.org/index.php/AAAI/article/download/16267/16074
[ "Steffen Jung", "Margret Keuper" ]
Recent advances in deep generative models for photo-realistic images have led to high quality visual results. Such models learn to generate data from a given training distribution such that generated images can not be easily distinguished from real images by the human eye. Yet, recent work on the detection of such fake...
main
Computer Vision
10.1609/aaai.v35i2.16267
35
2
1734-1742
official
2012.03110
title_snapshot
10.1609/aaai.v35i2.16268
StarNet: towards Weakly Supervised Few-Shot Object Detection
https://ojs.aaai.org/index.php/AAAI/article/view/16268
https://ojs.aaai.org/index.php/AAAI/article/download/16268/16075
[ "Leonid Karlinsky", "Joseph Shtok", "Amit Alfassy", "Moshe Lichtenstein", "Sivan Harary", "Eli Schwartz", "Sivan Doveh", "Prasanna Sattigeri", "Rogerio Feris", "Alex Bronstein", "Raja Giryes" ]
Few-shot detection and classification have advanced significantly in recent years. Yet, detection approaches require strong annotation (bounding boxes) both for pre-training and for adaptation to novel classes, and classification approaches rarely provide localization of objects in the scene. In this paper, we introduc...
main
Computer Vision
10.1609/aaai.v35i2.16268
35
2
1743-1753
official
2003.06798
title_snapshot
10.1609/aaai.v35i2.16269
Discriminative Region Suppression for Weakly-Supervised Semantic Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16269
https://ojs.aaai.org/index.php/AAAI/article/download/16269/16076
[ "Beomyoung Kim", "Sangeun Han", "Junmo Kim" ]
Weakly-supervised semantic segmentation (WSSS) using image-level labels has recently attracted much attention for reducing annotation costs. Existing WSSS methods utilize localization maps from the classification network to generate pseudo segmentation labels. However, since localization maps obtained from the classifi...
main
Computer Vision
10.1609/aaai.v35i2.16269
35
2
1754-1761
official
2103.07246
title_snapshot
10.1609/aaai.v35i2.16270
Visual Comfort Aware-Reinforcement Learning for Depth Adjustment of Stereoscopic 3D Images
https://ojs.aaai.org/index.php/AAAI/article/view/16270
https://ojs.aaai.org/index.php/AAAI/article/download/16270/16077
[ "Hak Gu Kim", "Minho Park", "Sangmin Lee", "Seongyeop Kim", "Yong Man Ro" ]
Depth adjustment aims to enhance the visual experience of stereoscopic 3D (S3D) images, which accompanied with improving visual comfort and depth perception. For a human expert, the depth adjustment procedure is a sequence of iterative decision making. The human expert iteratively adjusted the depth until he is satisfi...
main
Computer Vision
10.1609/aaai.v35i2.16270
35
2
1762-1770
official
2104.06782
title_snapshot
10.1609/aaai.v35i2.16271
Dual Compositional Learning in Interactive Image Retrieval
https://ojs.aaai.org/index.php/AAAI/article/view/16271
https://ojs.aaai.org/index.php/AAAI/article/download/16271/16078
[ "Jongseok Kim", "Youngjae Yu", "Hoeseong Kim", "Gunhee Kim" ]
We present an approach named Dual Composition Network (DCNet) for interactive image retrieval that searches for the best target image for a natural language query and a reference image. To accomplish this task, existing methods have focused on learning a composite representation of the reference image and the text quer...
main
Computer Vision
10.1609/aaai.v35i2.16271
35
2
1771-1779
official
null
null
10.1609/aaai.v35i2.16272
End-to-End Differentiable Learning to HDR Image Synthesis for Multi-exposure Images
https://ojs.aaai.org/index.php/AAAI/article/view/16272
https://ojs.aaai.org/index.php/AAAI/article/download/16272/16079
[ "Junghee Kim", "Siyeong Lee", "Suk-Ju Kang" ]
Recently, high dynamic range (HDR) image reconstruction based on the multiple exposure stack from a given single exposure utilizes a deep learning framework to generate high-quality HDR images. These conventional networks focus on the exposure transfer task to reconstruct the multi-exposure stack. Therefore, they often...
main
Computer Vision
10.1609/aaai.v35i2.16272
35
2
1780-1788
official
2006.15833
title_snapshot
10.1609/aaai.v35i2.16273
Structured Co-reference Graph Attention for Video-grounded Dialogue
https://ojs.aaai.org/index.php/AAAI/article/view/16273
https://ojs.aaai.org/index.php/AAAI/article/download/16273/16080
[ "Junyeong Kim", "Sunjae Yoon", "Dahyun Kim", "Chang D. Yoo" ]
A video-grounded dialogue system referred to as the Structured Co-reference Graph Attention (SCGA) is presented for decoding the answer sequence to a question regarding a given video while keeping track of the dialogue context. Although recent efforts have made great strides in improving the quality of the response, pe...
main
Computer Vision
10.1609/aaai.v35i2.16273
35
2
1789-1797
official
2103.13361
title_snapshot
10.1609/aaai.v35i2.16243
Progressive One-shot Human Parsing
https://ojs.aaai.org/index.php/AAAI/article/view/16243
https://ojs.aaai.org/index.php/AAAI/article/download/16243/16050
[ "Haoyu He", "Jing Zhang", "Bhavani Thuraisingham", "Dacheng Tao" ]
Prior human parsing models are limited to parsing humans into classes pre-defined in the training data, which is not flexible to generalize to unseen classes, e.g., new clothing in fashion analysis. In this paper, we propose a new problem named one-shot human parsing (OSHP) that requires to parse human into an open set...
main
Computer Vision
10.1609/aaai.v35i2.16243
35
2
1522-1530
official
2012.11810
title_snapshot
10.1609/aaai.v35i2.16244
Consistent-Separable Feature Representation for Semantic Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16244
https://ojs.aaai.org/index.php/AAAI/article/download/16244/16051
[ "Xingjian He", "Jing Liu", "Jun Fu", "Xinxin Zhu", "Jinqiao Wang", "Hanqing Lu" ]
Cross-entropy loss combined with softmax is one of the most commonly used supervision components in most existing segmentation methods. The softmax loss is typically good at optimizing the inter-class difference, but not good at reducing the intra-class variation, which can be suboptimal for semantic segmentation task....
main
Computer Vision
10.1609/aaai.v35i2.16244
35
2
1531-1539
official
null
null
10.1609/aaai.v35i2.16245
Error-Aware Density Isomorphism Reconstruction for Unsupervised Cross-Domain Crowd Counting
https://ojs.aaai.org/index.php/AAAI/article/view/16245
https://ojs.aaai.org/index.php/AAAI/article/download/16245/16052
[ "Yuhang He", "Zhiheng Ma", "Xing Wei", "Xiaopeng Hong", "Wei Ke", "Yihong Gong" ]
This paper focuses on the unsupervised domain adaptation problem for video-based crowd counting, in which we use labeled data as source domain and unlabelled video data as target domain. It is challenging as there is a huge gap between the source and the target domain and no annotations of samples are available in the ...
main
Computer Vision
10.1609/aaai.v35i2.16245
35
2
1540-1548
official
null
null
10.1609/aaai.v35i2.16246
DropLoss for Long-Tail Instance Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16246
https://ojs.aaai.org/index.php/AAAI/article/download/16246/16053
[ "Ting-I Hsieh", "Esther Robb", "Hwann-Tzong Chen", "Jia-Bin Huang" ]
Long-tailed class distributions are prevalent among the practical applications of object detection and instance segmentation. Prior work in long-tail instance segmentation addresses the imbalance of losses between rare and frequent categories by reducing the penalty for a model incorrectly predicting a rare class label...
main
Computer Vision
10.1609/aaai.v35i2.16246
35
2
1549-1557
official
2104.06402
title_snapshot
10.1609/aaai.v35i2.16247
Hand-Model-Aware Sign Language Recognition
https://ojs.aaai.org/index.php/AAAI/article/view/16247
https://ojs.aaai.org/index.php/AAAI/article/download/16247/16054
[ "Hezhen Hu", "Wengang Zhou", "Houqiang Li" ]
Hand gestures play a dominant role in the expression of sign language. Current deep-learning based video sign language recognition (SLR) methods usually follow a data-driven paradigm under the supervision of the category label. However, those methods suffer limited interpretability and may encounter the overfitting iss...
main
Computer Vision
10.1609/aaai.v35i2.16247
35
2
1558-1566
official
null
null
10.1609/aaai.v35i2.16248
Stratified Rule-Aware Network for Abstract Visual Reasoning
https://ojs.aaai.org/index.php/AAAI/article/view/16248
https://ojs.aaai.org/index.php/AAAI/article/download/16248/16055
[ "Sheng Hu", "Yuqing Ma", "Xianglong Liu", "Yanlu Wei", "Shihao Bai" ]
reasoning refers to the ability to analyze information, discover rules at an intangible level, and solve problems in innovative ways. Raven's Progressive Matrices (RPM) test is typically used to examine the capability of abstract reasoning. The subject is asked to identify the correct choice from the answer set to fill...
main
Computer Vision
10.1609/aaai.v35i2.16248
35
2
1567-1574
official
2002.06838
title_snapshot
10.1609/aaai.v35i2.16249
VIVO: Visual Vocabulary Pre-Training for Novel Object Captioning
https://ojs.aaai.org/index.php/AAAI/article/view/16249
https://ojs.aaai.org/index.php/AAAI/article/download/16249/16056
[ "Xiaowei Hu", "Xi Yin", "Kevin Lin", "Lei Zhang", "Jianfeng Gao", "Lijuan Wang", "Zicheng Liu" ]
It is highly desirable yet challenging to generate image captions that can describe novel objects which are unseen in caption-labeled training data, a capability that is evaluated in the novel object captioning challenge (nocaps). In this challenge, no additional image-caption training data, other than COCO Captions, i...
main
Computer Vision
10.1609/aaai.v35i2.16249
35
2
1575-1583
official
2009.13682
title_snapshot
10.1609/aaai.v35i2.16250
Exploiting Relationship for Complex-scene Image Generation
https://ojs.aaai.org/index.php/AAAI/article/view/16250
https://ojs.aaai.org/index.php/AAAI/article/download/16250/16057
[ "Tianyu Hua", "Hongdong Zheng", "Yalong Bai", "Wei Zhang", "Xiao-Ping Zhang", "Tao Mei" ]
The significant progress on Generative Adversarial Networks (GANs) has facilitated realistic single-object image generation based on language input. However, complex-scene generation (with various interactions among multiple objects) still suffers from messy layouts and object distortions, due to diverse configurations...
main
Computer Vision
10.1609/aaai.v35i2.16250
35
2
1584-1592
official
2104.00356
title_snapshot
10.1609/aaai.v35i2.16251
Modeling Deep Learning Based Privacy Attacks on Physical Mail
https://ojs.aaai.org/index.php/AAAI/article/view/16251
https://ojs.aaai.org/index.php/AAAI/article/download/16251/16058
[ "Bingyao Huang", "Ruyi Lian", "Dimitris Samaras", "Haibin Ling" ]
Mail privacy protection aims to prevent unauthorized access to hidden content within an envelope since normal paper envelopes are not as safe as we think. In this paper, for the first time, we show that with a well designed deep learning model, the hidden content may be largely recovered without opening the envelope. W...
main
Computer Vision
10.1609/aaai.v35i2.16251
35
2
1593-1601
official
2012.11803
title_snapshot
10.1609/aaai.v35i2.16252
PTN: A Poisson Transfer Network for Semi-supervised Few-shot Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16252
https://ojs.aaai.org/index.php/AAAI/article/download/16252/16059
[ "Huaxi Huang", "Junjie Zhang", "Jian Zhang", "Qiang Wu", "Chang Xu" ]
The predicament in semi-supervised few-shot learning (SSFSL) is to maximize the value of the extra unlabeled data to boost the few-shot learner. In this paper, we propose a Poisson Transfer Network (PTN) to mine the unlabeled information for SSFSL from two aspects. First, the Poisson Merriman–Bence–Osher (MBO) model bu...
main
Computer Vision
10.1609/aaai.v35i2.16252
35
2
1602-1609
official
2012.10844
title_snapshot
10.1609/aaai.v35i2.16253
Text-Guided Graph Neural Networks for Referring 3D Instance Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16253
https://ojs.aaai.org/index.php/AAAI/article/download/16253/16060
[ "Pin-Hao Huang", "Han-Hung Lee", "Hwann-Tzong Chen", "Tyng-Luh Liu" ]
This paper addresses a new task called referring 3D instance segmentation, which aims to segment out the target instance in a 3D scene given a query sentence. Previous work on scene understanding has explored visual grounding with natural language guidance, yet the emphasis is mostly constrained on images and videos. W...
main
Computer Vision
10.1609/aaai.v35i2.16253
35
2
1610-1618
official
null
null
10.1609/aaai.v35i2.16254
Initiative Defense against Facial Manipulation
https://ojs.aaai.org/index.php/AAAI/article/view/16254
https://ojs.aaai.org/index.php/AAAI/article/download/16254/16061
[ "Qidong Huang", "Jie Zhang", "Wenbo Zhou", "Weiming Zhang", "Nenghai Yu" ]
Benefiting from the development of generative adversarial networks (GAN), facial manipulation has achieved significant progress in both academia and industry recently. It inspires an increasing number of entertainment applications but also incurs severe threats to individual privacy and even political security meanwhil...
main
Computer Vision
10.1609/aaai.v35i2.16254
35
2
1619-1627
official
2112.10098
title_snapshot
10.1609/aaai.v35i2.16255
SnapMix: Semantically Proportional Mixing for Augmenting Fine-grained Data
https://ojs.aaai.org/index.php/AAAI/article/view/16255
https://ojs.aaai.org/index.php/AAAI/article/download/16255/16062
[ "Shaoli Huang", "Xinchao Wang", "Dacheng Tao" ]
Data mixing augmentation has proved effective in training deep models. Recent methods mix labels mainly according to the mixture proportion of image pixels. Due to the major discriminative information of a fine-grained image usually resides in subtle regions, these methods tend to introduce heavy label noise in fine-gr...
main
Computer Vision
10.1609/aaai.v35i2.16255
35
2
1628-1636
official
2012.04846
title_snapshot
10.1609/aaai.v35i2.16256
A Hybrid Attention Mechanism for Weakly-Supervised Temporal Action Localization
https://ojs.aaai.org/index.php/AAAI/article/view/16256
https://ojs.aaai.org/index.php/AAAI/article/download/16256/16063
[ "Ashraful Islam", "Chengjiang Long", "Richard Radke" ]
Weakly supervised temporal action localization is a challenging vision task due to the absence of ground-truth temporal locations of actions in the training videos. With only video-level supervision during training, most existing methods rely on a Multiple Instance Learning (MIL) framework to predict the start and end ...
main
Computer Vision
10.1609/aaai.v35i2.16256
35
2
1637-1645
official
2101.00545
title_snapshot
10.1609/aaai.v35i2.16257
Context-Aware Graph Convolution Network for Target Re-identification
https://ojs.aaai.org/index.php/AAAI/article/view/16257
https://ojs.aaai.org/index.php/AAAI/article/download/16257/16064
[ "Deyi Ji", "Haoran Wang", "Hanzhe Hu", "Weihao Gan", "Wei Wu", "Junjie Yan" ]
Most existing re-identification methods focus on learning robust and discriminative features with deep convolution networks. However, many of them consider content similarity separately and fail to utilize the context information of the query and gallery sets, e.g. probe-gallery and gallery-gallery relations, thus hard...
main
Computer Vision
10.1609/aaai.v35i2.16257
35
2
1646-1654
official
2012.04298
title_snapshot
10.1609/aaai.v35i2.16258
Improving Image Captioning by Leveraging Intra- and Inter-layer Global Representation in Transformer Network
https://ojs.aaai.org/index.php/AAAI/article/view/16258
https://ojs.aaai.org/index.php/AAAI/article/download/16258/16065
[ "Jiayi Ji", "Yunpeng Luo", "Xiaoshuai Sun", "Fuhai Chen", "Gen Luo", "Yongjian Wu", "Yue Gao", "Rongrong Ji" ]
Transformer-based architectures have shown great success in image captioning, where object regions are encoded and then attended into the vectorial representations to guide the caption decoding. However, such vectorial representations only contain region-level information without considering the global information refl...
main
Computer Vision
10.1609/aaai.v35i2.16258
35
2
1655-1663
official
2012.07061
title_snapshot
10.1609/aaai.v35i2.16259
Frequency Consistent Adaptation for Real World Super Resolution
https://ojs.aaai.org/index.php/AAAI/article/view/16259
https://ojs.aaai.org/index.php/AAAI/article/download/16259/16066
[ "Xiaozhong Ji", "Guangpin Tao", "Yun Cao", "Ying Tai", "Tong Lu", "Chengjie Wang", "Jilin Li", "Feiyue Huang" ]
Recent deep-learning based Super-Resolution (SR) methods have achieved remarkable performance on images with known degradation. However, these methods always fail in real-world scene, since the Low-Resolution (LR) images after the ideal degradation (e.g., bicubic down-sampling) deviate from real source domain. The doma...
main
Computer Vision
10.1609/aaai.v35i2.16259
35
2
1664-1672
official
2012.10102
title_snapshot
10.1609/aaai.v35i2.16260
Matching on Sets: Conquer Occluded Person Re-identification Without Alignment
https://ojs.aaai.org/index.php/AAAI/article/view/16260
https://ojs.aaai.org/index.php/AAAI/article/download/16260/16067
[ "Mengxi Jia", "Xinhua Cheng", "Yunpeng Zhai", "Shijian Lu", "Siwei Ma", "Yonghong Tian", "Jian Zhang" ]
Occluded person re-identification (re-ID) is a challenging task as different human parts may become invisible in cluttered scenes, making it hard to match person images of different identities. Most existing methods address this challenge by aligning spatial features of body parts according to semantic information (e.g...
main
Computer Vision
10.1609/aaai.v35i2.16260
35
2
1673-1681
official
null
null
10.1609/aaai.v35i2.16261
GradingNet: Towards Providing Reliable Supervisions for Weakly Supervised Object Detection by Grading the Box Candidates
https://ojs.aaai.org/index.php/AAAI/article/view/16261
https://ojs.aaai.org/index.php/AAAI/article/download/16261/16068
[ "Qifei Jia", "Shikui Wei", "Tao Ruan", "Yufeng Zhao", "Yao Zhao" ]
Weakly-Supervised Object Detection (WSOD) aims at training a model with limited and coarse annotations for precisely locating the regions of objects. Existing works solve the WSOD problem by using a two-stage framework, i.e., generating candidate bounding boxes with weak supervision information and then refining them b...
main
Computer Vision
10.1609/aaai.v35i2.16261
35
2
1682-1690
official
null
null
10.1609/aaai.v35i2.16262
SSN3D: Self-Separated Network to Align Parts for 3D Convolution in Video Person Re-Identification
https://ojs.aaai.org/index.php/AAAI/article/view/16262
https://ojs.aaai.org/index.php/AAAI/article/download/16262/16069
[ "Xiaoke Jiang", "Yu Qiao", "Junjie Yan", "Qichen Li", "Wanrong Zheng", "Dapeng Chen" ]
Temporal appearance misalignment is a crucial problem in video person re-identification. The same part of person (e.g. head or hand) appearing on different locations in video sequence weakens its discriminative ability, especially when we apply standard temporal aggregation such as 3D convolution or LSTM. To address th...
main
Computer Vision
10.1609/aaai.v35i2.16262
35
2
1691-1699
official
null
null
10.1609/aaai.v35i2.16223
Learning Complex 3D Human Self-Contact
https://ojs.aaai.org/index.php/AAAI/article/view/16223
https://ojs.aaai.org/index.php/AAAI/article/download/16223/16030
[ "Mihai Fieraru", "Mihai Zanfir", "Elisabeta Oneata", "Alin-Ionut Popa", "Vlad Olaru", "Cristian Sminchisescu" ]
Monocular estimation of three dimensional human self-contact is fundamental for detailed scene analysis including body language understanding and behaviour modeling. Existing 3d reconstruction methods do not focus on body regions in self-contact and consequently recover configurations that are either far from each othe...
main
Computer Vision
10.1609/aaai.v35i2.16223
35
2
1343-1351
official
2012.10366
title_snapshot
10.1609/aaai.v35i2.16224
Rain Streak Removal via Dual Graph Convolutional Network
https://ojs.aaai.org/index.php/AAAI/article/view/16224
https://ojs.aaai.org/index.php/AAAI/article/download/16224/16031
[ "Xueyang Fu", "Qi Qi", "Zheng-Jun Zha", "Yurui Zhu", "Xinghao Ding" ]
Deep convolutional neural networks (CNNs) have become dominant in the single image de-raining area. However, most deep CNNs-based de-raining methods are designed by stacking vanilla convolutional layers, which can only be used to model local relations. Therefore, long-range contextual information is rarely considered f...
main
Computer Vision
10.1609/aaai.v35i2.16224
35
2
1352-1360
official
null
null
10.1609/aaai.v35i2.16225
CompFeat: Comprehensive Feature Aggregation for Video Instance Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16225
https://ojs.aaai.org/index.php/AAAI/article/download/16225/16032
[ "Yang Fu", "Linjie Yang", "Ding Liu", "Thomas S. Huang", "Humphrey Shi" ]
Video instance segmentation is a complex task in which we need to detect, segment, and track each object for any given video. Previous approaches only utilize single-frame features for the detection, segmentation, and tracking of objects and they suffer in the video scenario due to several distinct challenges such as m...
main
Computer Vision
10.1609/aaai.v35i2.16225
35
2
1361-1369
official
2012.03400
title_snapshot
10.1609/aaai.v35i2.16226
Deep Metric Learning with Self-Supervised Ranking
https://ojs.aaai.org/index.php/AAAI/article/view/16226
https://ojs.aaai.org/index.php/AAAI/article/download/16226/16033
[ "Zheren Fu", "Yan Li", "Zhendong Mao", "Quan Wang", "Yongdong Zhang" ]
Deep metric learning aims to learn a deep embedding space, where similar objects are pushed towards together and different objects are repelled against. Existing approaches typically use inter-class characteristics, e.g. class-level information or instance-level similarity, to obtain semantic relevance of data points a...
main
Computer Vision
10.1609/aaai.v35i2.16226
35
2
1370-1378
official
null
null
10.1609/aaai.v35i2.16227
A Systematic Evaluation of Object Detection Networks for Scientific Plots
https://ojs.aaai.org/index.php/AAAI/article/view/16227
https://ojs.aaai.org/index.php/AAAI/article/download/16227/16034
[ "Pritha Ganguly", "Nitesh S Methani", "Mitesh M. Khapra", "Pratyush Kumar" ]
Are existing object detection methods adequate for detecting text and visual elements in scientific plots which are arguably different than the objects found in natural images? To answer this question, we train and compare the accuracy of Fast/Faster R-CNN, SSD, YOLO and RetinaNet on the PlotQA dataset with over 220,00...
main
Computer Vision
10.1609/aaai.v35i2.16227
35
2
1379-1387
official
2007.02240
title_snapshot
10.1609/aaai.v35i2.16228
The Complexity of Object Association in Multiple Object Tracking
https://ojs.aaai.org/index.php/AAAI/article/view/16228
https://ojs.aaai.org/index.php/AAAI/article/download/16228/16035
[ "Robert Ganian", "Thekla Hamm", "Sebastian Ordyniak" ]
Object association, i.e., the identification of which observations correspond to the same object, is a central task for the area of multiple object tracking. Two prominent models capturing this task have been introduced in the literature: the Lifted Multicut model and the more recent Lifted Paths model. Here, we carry ...
main
Computer Vision
10.1609/aaai.v35i2.16228
35
2
1388-1396
official
null
null
10.1609/aaai.v35i2.16229
Learning Local Neighboring Structure for Robust 3D Shape Representation
https://ojs.aaai.org/index.php/AAAI/article/view/16229
https://ojs.aaai.org/index.php/AAAI/article/download/16229/16036
[ "Zhongpai Gao", "Junchi Yan", "Guangtao Zhai", "Juyong Zhang", "Yiyan Yang", "Xiaokang Yang" ]
Mesh is a powerful data structure for 3D shapes. Representation learning for 3D meshes is important in many computer vision and graphics applications. The recent success of convolutional neural networks (CNNs) for structured data (e.g., images) suggests the value of adapting insight from CNN for 3D shapes. However, 3D ...
main
Computer Vision
10.1609/aaai.v35i2.16229
35
2
1397-1405
official
2004.09995
title_snapshot
10.1609/aaai.v35i2.16230
Semantic-guided Reinforced Region Embedding for Generalized Zero-Shot Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16230
https://ojs.aaai.org/index.php/AAAI/article/download/16230/16037
[ "Jiannan Ge", "Hongtao Xie", "Shaobo Min", "Yongdong Zhang" ]
Generalized zero-shot Learning (GZSL) aims to recognize images from either seen or unseen domain, mainly by learning a joint embedding space to associate image features with the corresponding category descriptions. Recent methods have proved that localizing important object regions can effectively bridge the semantic-v...
main
Computer Vision
10.1609/aaai.v35i2.16230
35
2
1406-1414
official
null
null
10.1609/aaai.v35i2.16231
Dynamic Graph Representation Learning for Video Dialog via Multi-Modal Shuffled Transformers
https://ojs.aaai.org/index.php/AAAI/article/view/16231
https://ojs.aaai.org/index.php/AAAI/article/download/16231/16038
[ "Shijie Geng", "Peng Gao", "Moitreya Chatterjee", "Chiori Hori", "Jonathan Le Roux", "Yongfeng Zhang", "Hongsheng Li", "Anoop Cherian" ]
Given an input video, its associated audio, and a brief caption, the audio-visual scene aware dialog (AVSD) task requires an agent to indulge in a question-answer dialog with a human about the audio-visual content. This task thus poses a challenging multi-modal representation learning and reasoning scenario, advancemen...
main
Computer Vision
10.1609/aaai.v35i2.16231
35
2
1415-1423
official
2007.03848
title_snapshot
10.1609/aaai.v35i2.16232
Boundary-Aware Geometric Encoding for Semantic Segmentation of Point Clouds
https://ojs.aaai.org/index.php/AAAI/article/view/16232
https://ojs.aaai.org/index.php/AAAI/article/download/16232/16039
[ "Jingyu Gong", "Jiachen Xu", "Xin Tan", "Jie Zhou", "Yanyun Qu", "Yuan Xie", "Lizhuang Ma" ]
Boundary information plays a significant role in 2D image segmentation, while usually being ignored in 3D point cloud segmentation where ambiguous features might be generated in feature extraction, leading to misclassification in the transition area between two objects. In this paper, firstly, we propose a Boundary Pre...
main
Computer Vision
10.1609/aaai.v35i2.16232
35
2
1424-1432
official
2101.02381
title_snapshot
10.1609/aaai.v35i2.16233
Analogical Image Translation for Fog Generation
https://ojs.aaai.org/index.php/AAAI/article/view/16233
https://ojs.aaai.org/index.php/AAAI/article/download/16233/16040
[ "Rui Gong", "Dengxin Dai", "Yuhua Chen", "Wen Li", "Danda Pani Paudel", "Luc Van Gool" ]
Image-to-image translation is to map images from a given style to another given style. While exceptionally successful, current methods assume the availability of training images in both source and target domains, which does not always hold in practice. Inspired by humans' reasoning capability of analogy, we propose ana...
main
Computer Vision
10.1609/aaai.v35i2.16233
35
2
1433-1441
official
2006.15618
title_snapshot
10.1609/aaai.v35i2.16234
Temporal ROI Align for Video Object Recognition
https://ojs.aaai.org/index.php/AAAI/article/view/16234
https://ojs.aaai.org/index.php/AAAI/article/download/16234/16041
[ "Tao Gong", "Kai Chen", "Xinjiang Wang", "Qi Chu", "Feng Zhu", "Dahua Lin", "Nenghai Yu", "Huamin Feng" ]
Video object detection is challenging in the presence of appearance deterioration in certain video frames. Therefore, it is a natural choice to aggregate temporal information from other frames of the same video into the current frame. However, ROI Align, as one of the most core procedures of video detectors, still rema...
main
Computer Vision
10.1609/aaai.v35i2.16234
35
2
1442-1450
official
2109.03495
title_snapshot
10.1609/aaai.v35i2.16235
SMART Frame Selection for Action Recognition
https://ojs.aaai.org/index.php/AAAI/article/view/16235
https://ojs.aaai.org/index.php/AAAI/article/download/16235/16042
[ "Shreyank N Gowda", "Marcus Rohrbach", "Laura Sevilla-Lara" ]
Video classification is computationally expensive. In this paper, we address theproblem of frame selection to reduce the computational cost of video classification.Recent work has successfully leveraged frame selection for long, untrimmed videos,where much of the content is not relevant, and easy to discard. In this wo...
main
Computer Vision
10.1609/aaai.v35i2.16235
35
2
1451-1459
official
2012.10671
title_snapshot
10.1609/aaai.v35i2.16236
Proxy Synthesis: Learning with Synthetic Classes for Deep Metric Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16236
https://ojs.aaai.org/index.php/AAAI/article/download/16236/16043
[ "Geonmo Gu", "Byungsoo Ko", "Han-Gyu Kim" ]
One of the main purposes of deep metric learning is to construct an embedding space that has well-generalized embeddings on both seen (training) classes and unseen (test) classes. Most existing works have tried to achieve this using different types of metric objectives and hard sample mining strategies with given train...
main
Computer Vision
10.1609/aaai.v35i2.16236
35
2
1460-1468
official
2103.15454
title_snapshot
10.1609/aaai.v35i2.16237
Interpretable Graph Capsule Networks for Object Recognition
https://ojs.aaai.org/index.php/AAAI/article/view/16237
https://ojs.aaai.org/index.php/AAAI/article/download/16237/16044
[ "Jindong Gu" ]
Capsule Networks, as alternatives to Convolutional Neural Networks, have been proposed to recognize objects from images. The current literature demonstrates many advantages of CapsNets over CNNs. However, how to create explanations for individual classifications of CapsNets has not been well explored. The widely used s...
main
Computer Vision
10.1609/aaai.v35i2.16237
35
2
1469-1477
official
2012.01674
title_snapshot
10.1609/aaai.v35i2.16238
Class-Incremental Instance Segmentation via Multi-Teacher Networks
https://ojs.aaai.org/index.php/AAAI/article/view/16238
https://ojs.aaai.org/index.php/AAAI/article/download/16238/16045
[ "Yanan Gu", "Cheng Deng", "Kun Wei" ]
Although deep neural networks have achieved amazing results on instance segmentation, they are still ill-equipped when they are required to learn new tasks incrementally. Concretely, they suffer from “catastrophic forgetting”, an abrupt degradation of performance on old classes with the initial training data missing. M...
main
Computer Vision
10.1609/aaai.v35i2.16238
35
2
1478-1486
official
null
null
10.1609/aaai.v35i2.16239
EfficientDeRain: Learning Pixel-wise Dilation Filtering for High-Efficiency Single-Image Deraining
https://ojs.aaai.org/index.php/AAAI/article/view/16239
https://ojs.aaai.org/index.php/AAAI/article/download/16239/16046
[ "Qing Guo", "Jingyang Sun", "Felix Juefei-Xu", "Lei Ma", "Xiaofei Xie", "Wei Feng", "Yang Liu", "Jianjun Zhao" ]
Single-image deraining is rather challenging due to the unknown rain model. Existing methods often make specific assumptions of the rain model, which can hardly cover many diverse circumstances in the real world, compelling them to employ complex optimization or progressive refinement. This, however, significantly affe...
main
Computer Vision
10.1609/aaai.v35i2.16239
35
2
1487-1495
official
2009.09238
title_snapshot
10.1609/aaai.v35i2.16240
Order Regularization on Ordinal Loss for Head Pose, Age and Gaze Estimation
https://ojs.aaai.org/index.php/AAAI/article/view/16240
https://ojs.aaai.org/index.php/AAAI/article/download/16240/16047
[ "Tianchu Guo", "Hui Zhang", "ByungIn Yoo", "Yongchao Liu", "Youngjun Kwak", "Jae-Joon Han" ]
Ordinal loss is widely used in solving regression problems with deep learning technologies. Its basic idea is to convert regression to classification while preserving the natural order. However, the order constraint is enforced only by ordinal label implicitly, leading to the real output values not strictly in order. I...
main
Computer Vision
10.1609/aaai.v35i2.16240
35
2
1496-1504
official
null
null
10.1609/aaai.v35i2.16241
Decoupled and Memory-Reinforced Networks: Towards Effective Feature Learning for One-Step Person Search
https://ojs.aaai.org/index.php/AAAI/article/view/16241
https://ojs.aaai.org/index.php/AAAI/article/download/16241/16048
[ "Chuchu Han", "Zhedong Zheng", "Changxin Gao", "Nong Sang", "Yi Yang" ]
The goal of person search is to localize and match query persons from scene images. For high efficiency, one-step methods have been developed to jointly handle the pedestrian detection and identification sub-tasks using a single network. There are two major challenges in the current one-step approaches. One is the mutu...
main
Computer Vision
10.1609/aaai.v35i2.16241
35
2
1505-1512
official
2102.10795
title_snapshot
10.1609/aaai.v35i2.16242
Spherical Image Generation from a Single Image by Considering Scene Symmetry
https://ojs.aaai.org/index.php/AAAI/article/view/16242
https://ojs.aaai.org/index.php/AAAI/article/download/16242/16049
[ "Takayuki Hara", "Yusuke Mukuta", "Tatsuya Harada" ]
Spherical images taken in all directions (360 degrees by 180 degrees) allow the full surroundings of a subject to be represented, providing an immersive experience to viewers. Generating a spherical image from a single normal-field-of-view (NFOV) image is convenient and expands the usage scenarios considerably without ...
main
Computer Vision
10.1609/aaai.v35i2.16242
35
2
1513-1521
official
null
null
10.1609/aaai.v35i2.16203
DramaQA: Character-Centered Video Story Understanding with Hierarchical QA
https://ojs.aaai.org/index.php/AAAI/article/view/16203
https://ojs.aaai.org/index.php/AAAI/article/download/16203/16010
[ "Seongho Choi", "Kyoung-Woon On", "Yu-Jung Heo", "Ahjeong Seo", "Youwon Jang", "Minsu Lee", "Byoung-Tak Zhang" ]
Despite recent progress on computer vision and natural language processing, developing a machine that can understand video story is still hard to achieve due to the intrinsic difficulty of video story. Moreover, researches on how to evaluate the degree of video understanding based on human cognitive process have not pr...
main
Computer Vision
10.1609/aaai.v35i2.16203
35
2
1166-1174
official
2005.03356
title_snapshot
10.1609/aaai.v35i2.16204
DeepCollaboration: Collaborative Generative and Discriminative Models for Class Incremental Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16204
https://ojs.aaai.org/index.php/AAAI/article/download/16204/16011
[ "Bo Cui", "Guyue Hu", "Shan Yu" ]
An important challenge for neural networks is to learn incrementally, i.e., learn new classes without catastrophic forgetting. To overcome this problem, generative replay technique has been suggested, which can generate samples belonging to learned classes while learning new ones. However, such generative models usuall...
main
Computer Vision
10.1609/aaai.v35i2.16204
35
2
1175-1183
official
null
null
10.1609/aaai.v35i2.16205
Split then Refine: Stacked Attention-guided ResUNets for Blind Single Image Visible Watermark Removal
https://ojs.aaai.org/index.php/AAAI/article/view/16205
https://ojs.aaai.org/index.php/AAAI/article/download/16205/16012
[ "Xiaodong Cun", "Chi-Man Pun" ]
Digital watermark is a commonly used technique to protect the copyright of medias. Simultaneously, to increase the robustness of watermark, attacking technique, such as watermark removal, also gets the attention from the community. Previous watermark removal methods require to gain the watermark location from users or ...
main
Computer Vision
10.1609/aaai.v35i2.16205
35
2
1184-1192
official
2012.07007
title_snapshot
10.1609/aaai.v35i2.16206
RSGNet: Relation based Skeleton Graph Network for Crowded Scenes Pose Estimation
https://ojs.aaai.org/index.php/AAAI/article/view/16206
https://ojs.aaai.org/index.php/AAAI/article/download/16206/16013
[ "Yan Dai", "Xuanhan Wang", "Lianli Gao", "Jingkuan Song", "Heng Tao Shen" ]
Despite of the recent great progress on multi-person pose estimation, existing solutions still remain challenging under the condition of "crowded scenes'', where RGB images capture complex real-world scenes with highly-overlapped people, severe occlusions and diverse postures. In this work, we focus on two main problem...
main
Computer Vision
10.1609/aaai.v35i2.16206
35
2
1193-1200
official
null
null
10.1609/aaai.v35i2.16207
Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detection
https://ojs.aaai.org/index.php/AAAI/article/view/16207
https://ojs.aaai.org/index.php/AAAI/article/download/16207/16014
[ "Jiajun Deng", "Shaoshuai Shi", "Peiwei Li", "Wengang Zhou", "Yanyong Zhang", "Houqiang Li" ]
Recent advances on 3D object detection heavily rely on how the 3D data are represented, i.e., voxel-based or point-based representation. Many existing high performance 3D detectors are point-based because this structure can better retain precise point positions. Nevertheless, point-level features lead to high computati...
main
Computer Vision
10.1609/aaai.v35i2.16207
35
2
1201-1209
official
2012.15712
title_snapshot
10.1609/aaai.v35i2.16208
Arbitrary Video Style Transfer via Multi-Channel Correlation
https://ojs.aaai.org/index.php/AAAI/article/view/16208
https://ojs.aaai.org/index.php/AAAI/article/download/16208/16015
[ "Yingying Deng", "Fan Tang", "Weiming Dong", "Haibin Huang", "Chongyang Ma", "Changsheng Xu" ]
Video style transfer is attracting increasing attention from the artificial intelligence community because of its numerous applications, such as augmented reality and animation production. Relative to traditional image style transfer, video style transfer presents new challenges, including how to effectively generate s...
main
Computer Vision
10.1609/aaai.v35i2.16208
35
2
1210-1217
official
2009.08003
title_snapshot
10.1609/aaai.v35i2.16209
Similarity Reasoning and Filtration for Image-Text Matching
https://ojs.aaai.org/index.php/AAAI/article/view/16209
https://ojs.aaai.org/index.php/AAAI/article/download/16209/16016
[ "Haiwen Diao", "Ying Zhang", "Lin Ma", "Huchuan Lu" ]
Image-text matching plays a critical role in bridging the vision and language, and great progress has been made by exploiting the global alignment between image and sentence, or local alignments between regions and words. However, how to make the most of these alignments to infer more accurate matching scores is still ...
main
Computer Vision
10.1609/aaai.v35i2.16209
35
2
1218-1226
official
2101.01368
title_snapshot
10.1609/aaai.v35i2.16210
Spatio-Temporal Difference Descriptor for Skeleton-Based Action Recognition
https://ojs.aaai.org/index.php/AAAI/article/view/16210
https://ojs.aaai.org/index.php/AAAI/article/download/16210/16017
[ "Chongyang Ding", "Kai Liu", "Jari Korhonen", "Evgeny Belyaev" ]
In skeletal representation, intra-frame differences between body joints, as well as inter-frame dynamics between body skeletons contain discriminative information for action recognition. Conventional methods for modeling human skeleton sequences generally depend on motion trajectory and body joint dependency informatio...
main
Computer Vision
10.1609/aaai.v35i2.16210
35
2
1227-1235
official
null
null
10.1609/aaai.v35i2.16211
Towards Universal Physical Attacks on Single Object Tracking
https://ojs.aaai.org/index.php/AAAI/article/view/16211
https://ojs.aaai.org/index.php/AAAI/article/download/16211/16018
[ "Li Ding", "Yongwei Wang", "Kaiwen Yuan", "Minyang Jiang", "Ping Wang", "Hua Huang", "Z. Jane Wang" ]
Recent studies show that small perturbations in video frames could misguide single object trackers. However, such attacks have been mainly designed for digital-domain videos (i.e., perturbation on full images), which makes them practically infeasible to evaluate the adversarial vulnerability of trackers in real-world s...
main
Computer Vision
10.1609/aaai.v35i2.16211
35
2
1236-1245
official
null
null
10.1609/aaai.v35i2.16212
Modeling the Probabilistic Distribution of Unlabeled Data for One-shot Medical Image Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16212
https://ojs.aaai.org/index.php/AAAI/article/download/16212/16019
[ "Yuhang Ding", "Xin Yu", "Yi Yang" ]
Existing image segmentation networks mainly leverage large-scale labeled datasets to attain high accuracy. However, labeling medical images is very expensive since it requires sophisticated expert knowledge. Thus, it is more desirable to employ only a few labeled data in pursuing high segmentation performance. In this ...
main
Computer Vision
10.1609/aaai.v35i2.16212
35
2
1246-1254
official
2102.02033
title_judge
10.1609/aaai.v35i2.16213
Few-Shot Class-Incremental Learning via Relation Knowledge Distillation
https://ojs.aaai.org/index.php/AAAI/article/view/16213
https://ojs.aaai.org/index.php/AAAI/article/download/16213/16020
[ "Songlin Dong", "Xiaopeng Hong", "Xiaoyu Tao", "Xinyuan Chang", "Xing Wei", "Yihong Gong" ]
In this paper, we focus on the challenging few-shot class incremental learning (FSCIL) problem, which requires to transfer knowledge from old tasks to new ones and solves catastrophic forgetting. We propose the exemplar relation distillation incremental learning framework to balance the tasks of old-knowledge preservin...
main
Computer Vision
10.1609/aaai.v35i2.16213
35
2
1255-1263
official
null
null
10.1609/aaai.v35i2.16214
MIEHDR CNN: Main Image Enhancement based Ghost-Free High Dynamic Range Imaging using Dual-Lens Systems
https://ojs.aaai.org/index.php/AAAI/article/view/16214
https://ojs.aaai.org/index.php/AAAI/article/download/16214/16021
[ "Xuan Dong", "Xiaoyan Hu", "Weixin Li", "Xiaojie Wang", "Yunhong Wang" ]
We study the High Dynamic Range (HDR) imaging problem using two Low Dynamic Range (LDR) images that are shot from dual-lens systems in a single shot time with different exposures. In most of the related HDR imaging methods, the problem is usually solved by Multiple Images Merging, i.e. the final HDR image is fused from...
main
Computer Vision
10.1609/aaai.v35i2.16214
35
2
1264-1272
official
null
null
10.1609/aaai.v35i2.16215
Boosting Image-based Mutual Gaze Detection using Pseudo 3D Gaze
https://ojs.aaai.org/index.php/AAAI/article/view/16215
https://ojs.aaai.org/index.php/AAAI/article/download/16215/16022
[ "Bardia Doosti", "Ching-Hui Chen", "Raviteja Vemulapalli", "Xuhui Jia", "Yukun Zhu", "Bradley Green" ]
Mutual gaze detection, i.e., predicting whether or not two people are looking at each other, plays an important role in understanding human interactions. In this work, we focus on the task of image-based mutual gaze detection, and propose a simple and effective approach to boost the performance by using an auxiliary 3D...
main
Computer Vision
10.1609/aaai.v35i2.16215
35
2
1273-1281
official
2010.07811
title_snapshot
10.1609/aaai.v35i2.16216
How to Save your Annotation Cost for Panoptic Segmentation?
https://ojs.aaai.org/index.php/AAAI/article/view/16216
https://ojs.aaai.org/index.php/AAAI/article/download/16216/16023
[ "Xuefeng Du", "ChenHan Jiang", "Hang Xu", "Gengwei Zhang", "Zhenguo Li" ]
How to properly reduce the annotation cost for panoptic segmentation? How to leverage and optimize the cost-quality trade-off for training data and model? These questions are key challenges towards a label-efficient and scalable panoptic segmentation system due to its expensive instance/semantic pixel-level annotation ...
main
Computer Vision
10.1609/aaai.v35i2.16216
35
2
1282-1290
official
null
null
10.1609/aaai.v35i2.16217
DIRV: Dense Interaction Region Voting for End-to-End Human-Object Interaction Detection
https://ojs.aaai.org/index.php/AAAI/article/view/16217
https://ojs.aaai.org/index.php/AAAI/article/download/16217/16024
[ "Hao-Shu Fang", "Yichen Xie", "Dian Shao", "Cewu Lu" ]
Recent years, human-object interaction (HOI) detection has achieved impressive advances. However, conventional two-stage methods are usually slow in inference. On the other hand, existing one-stage methods mainly focus on the union regions of interactions, which introduce unnecessary visual information as disturbances ...
main
Computer Vision
10.1609/aaai.v35i2.16217
35
2
1291-1299
official
2010.01005
title_snapshot
10.1609/aaai.v35i2.16218
DecAug: Augmenting HOI Detection via Decomposition
https://ojs.aaai.org/index.php/AAAI/article/view/16218
https://ojs.aaai.org/index.php/AAAI/article/download/16218/16025
[ "Hao-Shu Fang", "Yichen Xie", "Dian Shao", "Yong-Lu Li", "Cewu Lu" ]
Human-object interaction (HOI) detection requires a large amount of annotated data. Current algorithms suffer from insufficient training samples and category imbalance within datasets. To increase data efficiency, in this paper, we propose an efficient and effective data augmentation method called DecAug for HOI detect...
main
Computer Vision
10.1609/aaai.v35i2.16218
35
2
1300-1308
official
2010.01007
title_snapshot
10.1609/aaai.v35i2.16219
Partially Non-Autoregressive Image Captioning
https://ojs.aaai.org/index.php/AAAI/article/view/16219
https://ojs.aaai.org/index.php/AAAI/article/download/16219/16026
[ "Zhengcong Fei" ]
Current state-of-the-art image captioning systems usually generated descriptions autoregressively, i.e., every forward step conditions on the given image and previously produced words. The sequential attribution causes a unavoidable decoding latency. Non-autoregressive image captioning, on the other hand, predicts the ...
main
Computer Vision
10.1609/aaai.v35i2.16219
35
2
1309-1316
official
null
null
10.1609/aaai.v35i2.16220
Memory-Augmented Image Captioning
https://ojs.aaai.org/index.php/AAAI/article/view/16220
https://ojs.aaai.org/index.php/AAAI/article/download/16220/16027
[ "Zhengcong Fei" ]
Current deep learning-based image captioning systems have been proven to store practical knowledge with their parameters and achieve competitive performances in the public datasets. Nevertheless, their ability to access and precisely manipulate the mastered knowledge is still limited. Besides, providing evidence for de...
main
Computer Vision
10.1609/aaai.v35i2.16220
35
2
1317-1324
official
null
null
10.1609/aaai.v35i2.16221
Edge-competing Pathological Liver Vessel Segmentation with Limited Labels
https://ojs.aaai.org/index.php/AAAI/article/view/16221
https://ojs.aaai.org/index.php/AAAI/article/download/16221/16028
[ "Zunlei Feng", "Zhonghua Wang", "Xinchao Wang", "Xiuming Zhang", "Lechao Cheng", "Jie Lei", "Yuexuan Wang", "Mingli Song" ]
The microvascular invasion (MVI) is a major prognostic factor in hepatocellular carcinoma, which is one of the malignant tumors with the highest mortality rate. The diagnosis of MVI needs discovering the vessels that contain hepatocellular carcinoma cells and counting their number in each vessel, which depends heavily ...
main
Computer Vision
10.1609/aaai.v35i2.16221
35
2
1325-1333
official
2108.00384
title_snapshot
10.1609/aaai.v35i2.16222
Visual Boundary Knowledge Translation for Foreground Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16222
https://ojs.aaai.org/index.php/AAAI/article/download/16222/16029
[ "Zunlei Feng", "Lechao Cheng", "Xinchao Wang", "Xiang Wang", "Ya Jie Liu", "Xiangtong Du", "Mingli Song" ]
When confronted with objects of unknown types in an image, humans can effortlessly and precisely tell their visual boundaries. This recognition mechanism and underlying generalization capability seem to contrast to state-of-the-art image segmentation networks that rely on large-scale category-aware annotated training s...
main
Computer Vision
10.1609/aaai.v35i2.16222
35
2
1334-1342
official
2108.00379
title_snapshot
10.1609/aaai.v35i2.16183
CNN Profiler on Polar Coordinate Images for Tropical Cyclone Structure Analysis
https://ojs.aaai.org/index.php/AAAI/article/view/16183
https://ojs.aaai.org/index.php/AAAI/article/download/16183/15990
[ "Boyo Chen", "Buo-Fu Chen", "Chun Min Hsiao" ]
Convolutional neural networks (CNN) have achieved great success in analyzing tropical cyclones (TC) with satellite images in several tasks, such as TC intensity estimation. In contrast, TC structure, which is conventionally described by a few parameters estimated subjectively by meteorology specialists, is still hard t...
main
Computer Vision
10.1609/aaai.v35i2.16183
35
2
991-998
official
2010.15158
title_snapshot
10.1609/aaai.v35i2.16184
Commonsense Knowledge Aware Concept Selection For Diverse and Informative Visual Storytelling
https://ojs.aaai.org/index.php/AAAI/article/view/16184
https://ojs.aaai.org/index.php/AAAI/article/download/16184/15991
[ "Hong Chen", "Yifei Huang", "Hiroya Takamura", "Hideki Nakayama" ]
Visual storytelling is a task of generating relevant and interesting stories for given image sequences. In this work we aim at increasing the diversity of the generated stories while preserving the informative content from the images. We propose to foster the diversity and informativeness of a generated story by using ...
main
Computer Vision
10.1609/aaai.v35i2.16184
35
2
999-1008
official
2102.02963
title_snapshot
10.1609/aaai.v35i2.16185
Attention-based Multi-Level Fusion Network for Light Field Depth Estimation
https://ojs.aaai.org/index.php/AAAI/article/view/16185
https://ojs.aaai.org/index.php/AAAI/article/download/16185/15992
[ "Jiaxin Chen", "Shuo Zhang", "Youfang Lin" ]
Depth estimation from Light Field (LF) images is a crucial basis for LF related applications. Since multiple views with abundant information are available, how to effectively fuse features of these views is a key point for accurate LF depth estimation. In this paper, we propose a novel attention-based multi-level fusio...
main
Computer Vision
10.1609/aaai.v35i2.16185
35
2
1009-1017
official
null
null
10.1609/aaai.v35i2.16186
Joint Demosaicking and Denoising in the Wild: The Case of Training Under Ground Truth Uncertainty
https://ojs.aaai.org/index.php/AAAI/article/view/16186
https://ojs.aaai.org/index.php/AAAI/article/download/16186/15993
[ "Jierun Chen", "Song Wen", "S.-H. Gary Chan" ]
Image demosaicking and denoising are the two key fundamental steps in digital camera pipelines, aiming to reconstruct clean color images from noisy luminance readings. In this paper, we propose and study Wild-JDD, a novel learning framework for joint demosaicking and denoising in the wild. In contrast to previous works...
main
Computer Vision
10.1609/aaai.v35i2.16186
35
2
1018-1026
official
2101.04442
title_snapshot
10.1609/aaai.v35i2.16187
Spatial-temporal Causal Inference for Partial Image-to-video Adaptation
https://ojs.aaai.org/index.php/AAAI/article/view/16187
https://ojs.aaai.org/index.php/AAAI/article/download/16187/15994
[ "Jin Chen", "Xinxiao Wu", "Yao Hu", "Jiebo Luo" ]
Image-to-video adaptation leverages off-the-shelf learned models in labeled images to help classification in unlabeled videos, thus alleviating the high computation overhead of training a video classifier from scratch. This task is very challenging since there exist two types of domain shifts between images and videos:...
main
Computer Vision
10.1609/aaai.v35i2.16187
35
2
1027-1035
official
null
null
10.1609/aaai.v35i2.16188
Ref-NMS: Breaking Proposal Bottlenecks in Two-Stage Referring Expression Grounding
https://ojs.aaai.org/index.php/AAAI/article/view/16188
https://ojs.aaai.org/index.php/AAAI/article/download/16188/15995
[ "Long Chen", "Wenbo Ma", "Jun Xiao", "Hanwang Zhang", "Shih-Fu Chang" ]
The prevailing framework for solving referring expression grounding is based on a two-stage process: 1) detecting proposals with an object detector and 2) grounding the referent to one of the proposals. Existing two-stage solutions mostly focus on the grounding step, which aims to align the expressions with the proposa...
main
Computer Vision
10.1609/aaai.v35i2.16188
35
2
1036-1044
official
2009.01449
title_snapshot
10.1609/aaai.v35i2.16189
RSPNet: Relative Speed Perception for Unsupervised Video Representation Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16189
https://ojs.aaai.org/index.php/AAAI/article/download/16189/15996
[ "Peihao Chen", "Deng Huang", "Dongliang He", "Xiang Long", "Runhao Zeng", "Shilei Wen", "Mingkui Tan", "Chuang Gan" ]
We study unsupervised video representation learning that seeks to learn both motion and appearance features from unlabeled video only, which can be reused for downstream tasks such as action recognition. This task, however, is extremely challenging due to 1) the highly complex spatial-temporal information in videos and...
main
Computer Vision
10.1609/aaai.v35i2.16189
35
2
1045-1053
official
2011.07949
title_snapshot
10.1609/aaai.v35i2.16190
Dual Distribution Alignment Network for Generalizable Person Re-Identification
https://ojs.aaai.org/index.php/AAAI/article/view/16190
https://ojs.aaai.org/index.php/AAAI/article/download/16190/15997
[ "Peixian Chen", "Pingyang Dai", "Jianzhuang Liu", "Feng Zheng", "Mingliang Xu", "Qi Tian", "Rongrong Ji" ]
Domain generalization (DG) offers a preferable real-world setting for Person Re-Identification (Re-ID), which trains a model using multiple source domain datasets and expects it to perform well in an unseen target domain without any model updating. Unfortunately, most DG approaches are designed explicitly for classific...
main
Computer Vision
10.1609/aaai.v35i2.16190
35
2
1054-1062
official
2007.13249
title_snapshot
10.1609/aaai.v35i2.16191
RGB-D Salient Object Detection via 3D Convolutional Neural Networks
https://ojs.aaai.org/index.php/AAAI/article/view/16191
https://ojs.aaai.org/index.php/AAAI/article/download/16191/15998
[ "Qian Chen", "Ze Liu", "Yi Zhang", "Keren Fu", "Qijun Zhao", "Hongwei Du" ]
RGB-D salient object detection (SOD) recently has attracted increasing research interest and many deep learning methods based on encoder-decoder architectures have emerged. However, most existing RGB-D SOD models conduct feature fusion either in the single encoder or the decoder stage, which hardly guarantees sufficien...
main
Computer Vision
10.1609/aaai.v35i2.16191
35
2
1063-1071
official
2101.10241
title_snapshot
10.1609/aaai.v35i2.16192
Mind-the-Gap! Unsupervised Domain Adaptation for Text-Video Retrieval
https://ojs.aaai.org/index.php/AAAI/article/view/16192
https://ojs.aaai.org/index.php/AAAI/article/download/16192/15999
[ "Qingchao Chen", "Yang Liu", "Samuel Albanie" ]
When can we expect a text-video retrieval system to work effectively on datasets that differ from its training domain? In this work, we investigate this question through the lens of unsupervised domain adaptation in which the objective is to match natural language queries and video content in the presence of domain shi...
main
Computer Vision
10.1609/aaai.v35i2.16192
35
2
1072-1080
official
null
null
10.1609/aaai.v35i2.16193
Local Relation Learning for Face Forgery Detection
https://ojs.aaai.org/index.php/AAAI/article/view/16193
https://ojs.aaai.org/index.php/AAAI/article/download/16193/16000
[ "Shen Chen", "Taiping Yao", "Yang Chen", "Shouhong Ding", "Jilin Li", "Rongrong Ji" ]
With the rapid development of facial manipulation techniques, face forgery has received considerable attention in digital media forensics due to security concerns. Most existing methods formulate face forgery detection as a classification problem and utilize binary labels or manipulated region masks as supervision. How...
main
Computer Vision
10.1609/aaai.v35i2.16193
35
2
1081-1088
official
2105.02577
title_snapshot
10.1609/aaai.v35i2.16194
Deductive Learning for Weakly-Supervised 3D Human Pose Estimation via Uncalibrated Cameras
https://ojs.aaai.org/index.php/AAAI/article/view/16194
https://ojs.aaai.org/index.php/AAAI/article/download/16194/16001
[ "Xipeng Chen", "Pengxu Wei", "Liang Lin" ]
Without prohibitive and laborious 3D annotations, weakly-supervised 3D human pose methods mainly employ the model regularization with geometric projection consistency or geometry estimation from multi-view images. Nevertheless, those approaches explicitly need known parameters of calibrated cameras, exhibiting a limite...
main
Computer Vision
10.1609/aaai.v35i2.16194
35
2
1089-1096
official
null
null
10.1609/aaai.v35i2.16195
A Unified Multi-Scenario Attacking Network for Visual Object Tracking
https://ojs.aaai.org/index.php/AAAI/article/view/16195
https://ojs.aaai.org/index.php/AAAI/article/download/16195/16002
[ "Xuesong Chen", "Canmiao Fu", "Feng Zheng", "Yong Zhao", "Hongsheng Li", "Ping Luo", "Guo-Jun Qi" ]
Existing methods of adversarial attacks successfully generate adversarial examples to confuse Deep Neural Networks (DNNs) of image classification and object detection, resulting in wrong predictions. However, these methods are difficult to attack models of video object tracking, because the tracking algorithms could ha...
main
Computer Vision
10.1609/aaai.v35i2.16195
35
2
1097-1104
official
null
null
10.1609/aaai.v35i2.16196
SSD-GAN: Measuring the Realness in the Spatial and Spectral Domains
https://ojs.aaai.org/index.php/AAAI/article/view/16196
https://ojs.aaai.org/index.php/AAAI/article/download/16196/16003
[ "Yuanqi Chen", "Ge Li", "Cece Jin", "Shan Liu", "Thomas Li" ]
This paper observes that there is an issue of high frequencies missing in the discriminator of standard GAN, and we reveal it stems from downsampling layers employed in the network architecture. This issue makes the generator lack the incentive from the discriminator to learn high-frequency content of data, resulting i...
main
Computer Vision
10.1609/aaai.v35i2.16196
35
2
1105-1112
official
2012.05535
title_snapshot
10.1609/aaai.v35i2.16197
Multi-Scale Spatial Temporal Graph Convolutional Network for Skeleton-Based Action Recognition
https://ojs.aaai.org/index.php/AAAI/article/view/16197
https://ojs.aaai.org/index.php/AAAI/article/download/16197/16004
[ "Zhan Chen", "Sicheng Li", "Bing Yang", "Qinghan Li", "Hong Liu" ]
Graph convolutional networks have been widely used for skeleton-based action recognition due to their excellent modeling ability of non-Euclidean data. As the graph convolution is a local operation, it can only utilize the short-range joint dependencies and short-term trajectory but fails to directly model the distant ...
main
Computer Vision
10.1609/aaai.v35i2.16197
35
2
1113-1122
official
2206.13028
title_snapshot
10.1609/aaai.v35i2.16198
Cascade Network with Guided Loss and Hybrid Attention for Finding Good Correspondences
https://ojs.aaai.org/index.php/AAAI/article/view/16198
https://ojs.aaai.org/index.php/AAAI/article/download/16198/16005
[ "Zhi Chen", "Fan Yang", "Wenbing Tao" ]
Finding good correspondences is a critical prerequisite in many feature based tasks. Given a putative correspondence set of an image pair, we propose a neural network which finds correct correspondences by a binary-class classifier and estimates relative pose through classified correspondences. First, we analyze that d...
main
Computer Vision
10.1609/aaai.v35i2.16198
35
2
1123-1131
official
2102.00411
title_snapshot
10.1609/aaai.v35i2.16199
Generalizable Representation Learning for Mixture Domain Face Anti-Spoofing
https://ojs.aaai.org/index.php/AAAI/article/view/16199
https://ojs.aaai.org/index.php/AAAI/article/download/16199/16006
[ "Zhihong Chen", "Taiping Yao", "Kekai Sheng", "Shouhong Ding", "Ying Tai", "Jilin Li", "Feiyue Huang", "Xinyu Jin" ]
Face anti-spoofing approach based on domain generalization (DG) has drawn growing attention due to its robustness for unseen scenarios. Existing DG methods assume that the domain label is known. However, in real-world applications, the collected dataset always contains mixture domains, where the domain label is unknown...
main
Computer Vision
10.1609/aaai.v35i2.16199
35
2
1132-1139
official
2105.02453
title_snapshot
10.1609/aaai.v35i2.16200
SSPC-Net: Semi-supervised Semantic 3D Point Cloud Segmentation Network
https://ojs.aaai.org/index.php/AAAI/article/view/16200
https://ojs.aaai.org/index.php/AAAI/article/download/16200/16007
[ "Mingmei Cheng", "Le Hui", "Jin Xie", "Jian Yang" ]
Point cloud semantic segmentation is a crucial task in 3D scene understanding. Existing methods mainly focus on employing a large number of annotated labels for supervised semantic segmentation. Nonetheless, manually labeling such large point clouds for the supervised segmentation task is time-consuming. In order to re...
main
Computer Vision
10.1609/aaai.v35i2.16200
35
2
1140-1147
official
2104.07861
title_snapshot
10.1609/aaai.v35i2.16201
Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification
https://ojs.aaai.org/index.php/AAAI/article/view/16201
https://ojs.aaai.org/index.php/AAAI/article/download/16201/16008
[ "Siyuan Cheng", "Yingqi Liu", "Shiqing Ma", "Xiangyu Zhang" ]
Trojan (backdoor) attack is a form of adversarial attack on deep neural networks where the attacker provides victims with a model trained/retrained on malicious data. The backdoor can be activated when a normal input is stamped with a certain pattern called trigger, causing misclassification. Many existing trojan attac...
main
Computer Vision
10.1609/aaai.v35i2.16201
35
2
1148-1156
official
2012.11212
title_snapshot
10.1609/aaai.v35i2.16202
Graph and Temporal Convolutional Networks for 3D Multi-person Pose Estimation in Monocular Videos
https://ojs.aaai.org/index.php/AAAI/article/view/16202
https://ojs.aaai.org/index.php/AAAI/article/download/16202/16009
[ "Yu Cheng", "Bo Wang", "Bo Yang", "Robby T. Tan" ]
Despite the recent progress, 3D multi-person pose estimation from monocular videos is still challenging due to the commonly encountered problem of missing information caused by occlusion, partially out-of-frame target persons, and inaccurate person detection. To tackle this problem, we propose a novel framework integra...
main
Computer Vision
10.1609/aaai.v35i2.16202
35
2
1157-1165
official
2012.11806
title_snapshot
10.1609/aaai.v35i2.16170
Localization in the Crowd with Topological Constraints
https://ojs.aaai.org/index.php/AAAI/article/view/16170
https://ojs.aaai.org/index.php/AAAI/article/download/16170/15977
[ "Shahira Abousamra", "Minh Hoai", "Dimitris Samaras", "Chao Chen" ]
We address the problem of crowd localization, i.e., the prediction of dots corresponding to people in a crowded scene. Due to various challenges, a localization method is prone to spatial semantic errors, i.e., predicting multiple dots within a same person or collapsing multiple dots in a cluttered region. We propose a...
main
Computer Vision
10.1609/aaai.v35i2.16170
35
2
872-881
official
2012.12482
title_snapshot
10.1609/aaai.v35i2.16171
Deep Event Stereo Leveraged by Event-to-Image Translation
https://ojs.aaai.org/index.php/AAAI/article/view/16171
https://ojs.aaai.org/index.php/AAAI/article/download/16171/15978
[ "Soikat Hasan Ahmed", "Hae Woong Jang", "S M Nadim Uddin", "Yong Ju Jung" ]
Depth estimation in real-world applications requires precise responses to fast motion and challenging lighting conditions. Event cameras use bio-inspired event-driven sensors that provide instantaneous and asynchronous information of pixel-level log intensity changes, which makes them suitable for depth estimation in s...
main
Computer Vision
10.1609/aaai.v35i2.16171
35
2
882-890
official
null
null
10.1609/aaai.v35i2.16172
Optical Flow Estimation from a Single Motion-blurred Image
https://ojs.aaai.org/index.php/AAAI/article/view/16172
https://ojs.aaai.org/index.php/AAAI/article/download/16172/15979
[ "Dawit Mureja Argaw", "Junsik Kim", "Francois Rameau", "Jae Won Cho", "In So Kweon" ]
In most of computer vision applications, motion blur is regarded as an undesirable artifact. However, it has been shown that motion blur in an image may have practical interests in fundamental computer vision problems. In this work, we propose a novel framework to estimate optical flow from a single motion-blurred imag...
main
Computer Vision
10.1609/aaai.v35i2.16172
35
2
891-900
official
2103.02996
title_snapshot
10.1609/aaai.v35i2.16173
Motion-blurred Video Interpolation and Extrapolation
https://ojs.aaai.org/index.php/AAAI/article/view/16173
https://ojs.aaai.org/index.php/AAAI/article/download/16173/15980
[ "Dawit Mureja Argaw", "Junsik Kim", "Francois Rameau", "In So Kweon" ]
Abrupt motion of camera or objects in a scene result in a blurry video, and therefore recovering high quality video requires two types of enhancements: visual enhancement and temporal upsampling. A broad range of research attempted to recover clean frames from blurred image sequences or temporally upsample frames by in...
main
Computer Vision
10.1609/aaai.v35i2.16173
35
2
901-910
official
2103.02984
title_snapshot
10.1609/aaai.v35i2.16174
Disentangled Multi-Relational Graph Convolutional Network for Pedestrian Trajectory Prediction
https://ojs.aaai.org/index.php/AAAI/article/view/16174
https://ojs.aaai.org/index.php/AAAI/article/download/16174/15981
[ "Inhwan Bae", "Hae-Gon Jeon" ]
Pedestrian trajectory prediction is one of the important tasks required for autonomous navigation and social robots in human environments. Previous studies focused on estimating social forces among individual pedestrians. However, they did not consider the social forces of groups on pedestrians, which results in over-c...
main
Computer Vision
10.1609/aaai.v35i2.16174
35
2
911-919
official
null
null
10.1609/aaai.v35i2.16175
Dense Events Grounding in Video
https://ojs.aaai.org/index.php/AAAI/article/view/16175
https://ojs.aaai.org/index.php/AAAI/article/download/16175/15982
[ "Peijun Bao", "Qian Zheng", "Yadong Mu" ]
This paper explores a novel setting of temporal sentence grounding for the first time, dubbed as dense events grounding. Given an untrimmed video and a paragraph description, dense events grounding aims to jointly localize temporal moments of multiple events described in the paragraph. Our main motivating fact is that ...
main
Computer Vision
10.1609/aaai.v35i2.16175
35
2
920-928
official
null
null
10.1609/aaai.v35i2.16176
Context-aware Attentional Pooling (CAP) for Fine-grained Visual Classification
https://ojs.aaai.org/index.php/AAAI/article/view/16176
https://ojs.aaai.org/index.php/AAAI/article/download/16176/15983
[ "Ardhendu Behera", "Zachary Wharton", "Pradeep R P G Hewage", "Asish Bera" ]
Deep convolutional neural networks (CNNs) have shown a strong ability in mining discriminative object pose and parts information for image recognition. For fine-grained recognition, context-aware rich feature representation of object/scene plays a key role since it exhibits a significant variance in the same subcategor...
main
Computer Vision
10.1609/aaai.v35i2.16176
35
2
929-937
official
2101.06635
title_snapshot
10.1609/aaai.v35i2.16177
Appearance-Motion Memory Consistency Network for Video Anomaly Detection
https://ojs.aaai.org/index.php/AAAI/article/view/16177
https://ojs.aaai.org/index.php/AAAI/article/download/16177/15984
[ "Ruichu Cai", "Hao Zhang", "Wen Liu", "Shenghua Gao", "Zhifeng Hao" ]
Abnormal event detection in the surveillance video is an essential but challenging task, and many methods have been proposed to deal with this problem. The previous methods either only consider the appearance information or directly integrate the results of appearance and motion information without considering their en...
main
Computer Vision
10.1609/aaai.v35i2.16177
35
2
938-946
official
null
null
10.1609/aaai.v35i2.16178
Rethinking Object Detection in Retail Stores
https://ojs.aaai.org/index.php/AAAI/article/view/16178
https://ojs.aaai.org/index.php/AAAI/article/download/16178/15985
[ "Yuanqiang Cai", "Longyin Wen", "Libo Zhang", "Dawei Du", "Weiqiang Wang" ]
The conventional standard for object detection uses a bounding box to represent each individual object instance. However, it is not practical in the industry-relevant applications in the context of warehouses due to severe occlusions among groups of instances of the same categories. In this paper, we propose a new task...
main
Computer Vision
10.1609/aaai.v35i2.16178
35
2
947-954
official
2003.08230
title_snapshot
10.1609/aaai.v35i2.16179
YOLObile: Real-Time Object Detection on Mobile Devices via Compression-Compilation Co-Design
https://ojs.aaai.org/index.php/AAAI/article/view/16179
https://ojs.aaai.org/index.php/AAAI/article/download/16179/15986
[ "Yuxuan Cai", "Hongjia Li", "Geng Yuan", "Wei Niu", "Yanyu Li", "Xulong Tang", "Bin Ren", "Yanzhi Wang" ]
The rapid development and wide utilization of object detection techniques have aroused attention on both accuracy and speed of object detectors. However, the current state-of-the-art object detection works are either accuracy-oriented using a large model but leading to high latency or speed-oriented using a lightweight...
main
Computer Vision
10.1609/aaai.v35i2.16179
35
2
955-963
official
2009.05697
title_snapshot