Update README.md

README.md

@@ -11,18 +11,19 @@ tags:
 - 3d
 - fiftyone
 - image-classification
-dataset_summary: '
+dataset_summary: >
 
 
 
 
-  This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 1180 samples.
+  This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 1180
+  samples.
 
 
   ## Installation
 
 
-  If you haven''t already, install FiftyOne:
+  If you haven't already, install FiftyOne:
 
 
   ```bash
@@ -44,7 +45,7 @@ dataset_summary: '
 
   # Load the dataset
 
-  # Note: other available arguments include ''max_samples'', etc
+  # Note: other available arguments include 'max_samples', etc
 
   dataset = load_from_hub("harpreetsahota/crops3d")
 
@@ -54,8 +55,6 @@ dataset_summary: '
   session = fo.launch_app(dataset)
 
   ```
-
-  '
 ---
 
 # Dataset Card for crops3d
@@ -115,135 +114,191 @@ update_dataset_ply_paths(dataset)
 
 ```
 
+# Dataset Details
 
-## Dataset Details
+## Dataset Description
 
-### Dataset Description
+**Curated by:** Jianzhong Zhu, Ruifang Zhai, He Ren, Kai Xie, Aobo Du, Xinwei He, Chenxi Cui, Yinghua Wang, Junli Ye, Jiashi Wang, Xue Jiang, Yulong Wang, Chenglong Huang, Wanneng Yang (Huazhong Agricultural University, China)
 
-<!-- Provide a longer summary of what this dataset is. -->
+**Funded by [optional]:** National Natural Science Foundation of China, National Key Research and Development Program of China
 
+**Shared by [optional]:** Huazhong Agricultural University, National Key Laboratory of Crop Genetic Improvement
 
+**Language(s) (NLP):** Not applicable (3D point cloud dataset)
 
-- **Curated by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Language(s) (NLP):** en
-- **License:** [More Information Needed]
+**License:** Creative Commons Attribution 4.0 International (CC BY 4.0)
 
-### Dataset Sources [optional]
+## Dataset Sources [optional]
 
-<!-- Provide the basic links for the dataset. -->
+**Repository:** https://github.com/clawCa/Crops3D
 
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
+**Paper:** https://doi.org/10.1038/s41597-024-04290-0
 
-## Uses
+**Demo:** Available through FiftyOne visualization after processing with provided scripts
+
+You can see how the dataset was parsed [here](https://github.com/harpreetsahota204/crops3d_to_fiftyone)
 
-<!-- Address questions around how the dataset is intended to be used. -->
+## Uses
 
 ### Direct Use
 
-<!-- This section describes suitable use cases for the dataset. -->
+- **3D crop phenotyping research**: Analysis of plant morphology and structure
+
+- **Instance segmentation**: Segmenting individual plants in agricultural plots
+
+- **Plant classification**: Identifying and classifying different crop types from 3D data
 
-[More Information Needed]
+- **Organ segmentation**: Fine-grained segmentation of plant organs (leaves, stems, etc.)
 
-### Out-of-Scope Use
+- **Agricultural robotics**: Training perception models for autonomous agricultural systems
 
-<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
+- **Precision agriculture**: Development of crop monitoring and assessment tools
 
-[More Information Needed]
+- **Computer vision research**: Benchmarking 3D segmentation and classification algorithms
 
 ## Dataset Structure
 
-<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
+The dataset contains 1,180 high-quality 3D point cloud samples in PLY format with RGB color information:
+
+- **File format**: PLY (Polygon File Format) with XYZ coordinates and RGB colors
+
+- **Total samples**: 1,180 point clouds
+
+- **Crop types**: 8 (Maize: 225, Cabbage: 196, Cotton: 176, Rapeseed: 150, Wheat: 148, Potato: 118, Rice: 84, Tomato: 83)
+
+- **Acquisition methods**: Multi-view stereo (mvs) and structured light (sl)
+
+- **File naming**: `{CropType}-{identifier}.ply` (e.g., `Cabbage-mvs_1005_01.ply`)
 
-[More Information Needed]
+- **Point density**: Varies by acquisition method and crop complexity
+
+- **Additional variants**: 
+  - Crops3D_10k: Subsampled to 10,000 points using Farthest Point Sampling
+  - Crops3D_10k-C: Corruption robustness test sets with 7 corruption types at 5 severity levels
+  - Crops3D_IS: Instance segmentation data for plot-level analysis
 
 ## Dataset Creation
 
 ### Curation Rationale
 
-<!-- Motivation for the creation of this dataset. -->
-
-[More Information Needed]
+The dataset was created to address the lack of comprehensive 3D agricultural datasets for developing and benchmarking computer vision algorithms in precision agriculture. Existing datasets were limited in diversity, realism, and complexity, hindering the development of robust agricultural perception systems. Crops3D fills this gap by providing authentic, diverse, and complex 3D point cloud data from real agricultural environments.
 
 ### Source Data
 
-<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
-
 #### Data Collection and Processing
 
-<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
+- **Collection period**: Multiple growth stages across agricultural seasons
 
-[More Information Needed]
+- **Location**: Agricultural research fields in China
 
-#### Who are the source data producers?
+- **Equipment**: Professional 3D scanning devices using multi-view stereo and structured light technologies
 
-<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
+- **Processing**: 
+  - Raw point clouds captured from multiple viewpoints
+  - Registration and fusion to create complete 3D models
+  - Color information preserved from RGB cameras
+  - Quality control to ensure accurate geometry and color
 
-[More Information Needed]
+#### Who are the source data producers?
 
-### Annotations [optional]
+The data was produced by researchers at:
+- Huazhong Agricultural University
+- National Key Laboratory of Crop Genetic Improvement
+- Collaborative teams from agricultural and computer vision research groups
 
-<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
+### Annotations
 
 #### Annotation process
 
-<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
-
-[More Information Needed]
+- **Plant-level labels**: Each point cloud is labeled with its crop type
+- **Instance segmentation**: Plot-level data includes individual plant boundaries
+- **Organ segmentation**: Selected samples include organ-level annotations (leaves, stems)
+- **Quality assurance**: Multiple rounds of verification by agricultural experts
 
 #### Who are the annotators?
 
-<!-- This section describes the people or systems who created the annotations. -->
-
-[More Information Needed]
-
-#### Personal and Sensitive Information
+- Agricultural researchers and graduate students from Huazhong Agricultural University
+- Domain experts in crop phenotyping and plant science
+- Computer vision researchers familiar with 3D data annotation
 
-<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
+## Personal and Sensitive Information
 
-[More Information Needed]
+The dataset contains no personal or sensitive information. All data consists of 3D scans of agricultural crops in research fields with no human subjects or private information included.
 
 ## Bias, Risks, and Limitations
 
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+**Biases:**
+- **Geographic bias**: Data collected primarily from agricultural fields in China, may not represent all global agricultural conditions
+- **Crop selection bias**: Limited to 8 major crop types, doesn't cover all agricultural species
+- **Growth stage bias**: May not equally represent all growth stages for each crop
 
-[More Information Needed]
+**Risks:**
+- **Generalization**: Models trained on this data may not perform well on crops grown in significantly different climates or conditions
+- **Scale limitations**: Individual plant models may not scale directly to large field applications
 
-### Recommendations
+**Limitations:**
+- **Point cloud density variations**: Different acquisition methods produce varying point densities
+- **Occlusion**: Natural self-occlusion in dense crop canopies may limit visibility of internal structures
+- **Weather conditions**: Data collected under specific weather conditions may not represent all scenarios
+- **File size**: Large PLY files (~9GB total) require significant storage and processing resources
 
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+## Recommendations
 
-Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
+Users should be made aware of the following recommendations:
 
-## Citation [optional]
+1. **Preprocessing**: Consider subsampling large point clouds for efficiency (Crops3D_10k variant available)
+2. **Validation**: Test model generalization on crops from different geographic regions
+3. **Augmentation**: Apply appropriate 3D augmentations to improve model robustness
+4. **Multi-modal approaches**: Consider combining with 2D imagery or other sensor data for comprehensive analysis
+5. **Computational resources**: Ensure adequate GPU memory and storage for processing large point clouds
+6. **Domain adaptation**: May require fine-tuning for specific agricultural environments or crop varieties
 
-<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
+## Citation
 
 **BibTeX:**
 
-[More Information Needed]
+```bibtex
+@article{zhu2024crops3d,
+  title={Crops3D: a diverse 3D crop dataset for realistic perception and segmentation toward agricultural applications},
+  author={Zhu, Jianzhong and Zhai, Ruifang and Ren, He and Xie, Kai and Du, Aobo and He, Xinwei and Cui, Chenxi and Wang, Yinghua and Ye, Junli and Wang, Jiashi and Jiang, Xue and Wang, Yulong and Huang, Chenglong and Yang, Wanneng},
+  journal={Scientific Data},
+  volume={11},
+  number={1438},
+  year={2024},
+  doi={10.1038/s41597-024-04290-0},
+  publisher={Nature Publishing Group}
+}
+```
 
 **APA:**
 
-[More Information Needed]
-
-## Glossary [optional]
+Zhu, J., Zhai, R., Ren, H., Xie, K., Du, A., He, X., Cui, C., Wang, Y., Ye, J., Wang, J., Jiang, X., Wang, Y., Huang, C., & Yang, W. (2024). Crops3D: a diverse 3D crop dataset for realistic perception and segmentation toward agricultural applications. *Scientific Data*, 11, 1438. https://doi.org/10.1038/s41597-024-04290-0
 
-<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
+## Glossary
 
-[More Information Needed]
+- **PLY**: Polygon File Format, a file format for storing 3D data
+- **Point Cloud**: A collection of 3D points representing the external surface of an object
+- **FPS**: Farthest Point Sampling, a method for subsampling point clouds
+- **Instance Segmentation**: Identifying and separating individual objects (plants) in a scene
+- **Organ Segmentation**: Identifying different parts of a plant (leaves, stems, fruits)
+- **Multi-view Stereo (MVS)**: 3D reconstruction technique using multiple 2D images
+- **Structured Light (SL)**: 3D scanning method using projected light patterns
+- **Phenotyping**: Measuring observable plant characteristics
 
-## More Information [optional]
+## More Information
 
-[More Information Needed]
+- **Dataset website**: https://figshare.com/articles/dataset/Crops3D_a_diverse_3D_crop_dataset_for_realistic_perception_and_segmentation_toward_agricultural_applications/27313272
+- **Download link**: https://springernature.figshare.com/ndownloader/files/50027964
+- **Related datasets**: PlantNet, AgriNet, Leaf Counting Challenge datasets
+- **Applications**: Yield prediction, disease detection, growth monitoring, harvesting automation
 
-## Dataset Card Authors [optional]
+## Dataset Card Authors
 
-[More Information Needed]
+- Original dataset: Jianzhong Zhu et al., Huazhong Agricultural University
+- Dataset card: Prepared for Hugging Face distribution
 
 ## Dataset Card Contact
 
-[More Information Needed]
+- **Primary contact**: Wanneng Yang (yangwn@mail.hzau.edu.cn)
+- **Institution**: National Key Laboratory of Crop Genetic Improvement, Huazhong Agricultural University
+- **GitHub Issues**: https://github.com/clawCa/Crops3D/issues