Shifting AI Efficiency From Model-Centric to Data-Centric Compression
Paper
•
2505.19147
•
Published
•
144
Innovator
community
AI & ML interests
None defined yet.
linfengZ
authored
a
paper
8 months ago
linfengZ
authored
15
papers
10 months ago
Ada3D : Exploiting the Spatial Redundancy with Adaptive Inference for Efficient 3D Object Detection
Paper
•
2307.08209
•
Published
•
1
Finding the Task-Optimal Low-Bit Sub-Distribution in Deep Neural Networks
Paper
•
2112.15139
•
Published
PointDistiller: Structured Knowledge Distillation Towards Efficient and Compact 3D Detection
Paper
•
2205.11098
•
Published
Contrastive Deep Supervision
Paper
•
2207.05306
•
Published
Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?
Paper
•
2212.08320
•
Published
DiTFastAttn: Attention Compression for Diffusion Transformer Models
Paper
•
2406.08552
•
Published
•
25
Accelerating Diffusion Transformers with Token-wise Feature Caching
Paper
•
2410.05317
•
Published
SAVEn-Vid: Synergistic Audio-Visual Integration for Enhanced Understanding in Long Video Context
Paper
•
2411.16213
•
Published
•
2
Multi-Stage Vision Token Dropping: Towards Efficient Multimodal Large Language Model
Paper
•
2411.10803
•
Published
Compression with Global Guidance: Towards Training-free High-Resolution MLLMs Acceleration
Paper
•
2501.05179
•
Published
RealRAG: Retrieval-augmented Realistic Image Generation via Self-reflective Contrastive Learning
Paper
•
2502.00848
•
Published
•
1
Stop Looking for Important Tokens in Multimodal Language Models: Duplication Matters More
Paper
•
2502.11494
•
Published
ProReflow: Progressive Reflow with Decomposed Velocity
Paper
•
2503.04824
•
Published
•
9
Decouple-Then-Merge: Finetune Diffusion Models as Multi-Task Learning
Paper
•
2410.06664
•
Published
•
1
LEGION: Learning to Ground and Explain for Synthetic Image Detection
Paper
•
2503.15264
•
Published
•
21