Upload LookBench v20251201

README.md

@@ -1,3 +1,194 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+task_categories:
+  - image-feature-extraction
+  - zero-shot-image-classification
+language:
+  - en
+tags:
+  - fashion
+  - image-retrieval
+  - benchmark
+  - e-commerce
+  - visual-search
+pretty_name: LookBench
+size_categories:
+  - 10K<n<100K
+configs:
+  - config_name: aigen_streetlook
+    data_files:
+      - split: query
+        path: "v20251201/aigen_streetlook/query.parquet"
+      - split: gallery
+        path: "v20251201/aigen_streetlook/gallery.parquet"
+  - config_name: aigen_studio
+    data_files:
+      - split: query
+        path: "v20251201/aigen_studio/query.parquet"
+      - split: gallery
+        path: "v20251201/aigen_studio/gallery.parquet"
+  - config_name: real_streetlook
+    data_files:
+      - split: query
+        path: "v20251201/real_streetlook/query.parquet"
+      - split: gallery
+        path: "v20251201/real_streetlook/gallery.parquet"
+  - config_name: real_studio_flat
+    data_files:
+      - split: query
+        path: "v20251201/real_studio_flat/query.parquet"
+      - split: gallery
+        path: "v20251201/real_studio_flat/gallery.parquet"
+  - config_name: noise
+    data_files:
+      - split: gallery
+        path: "v20251201/noise/*.parquet"
+dataset_info:
+  features:
+    - name: image
+      dtype: image
+    - name: category
+      dtype: string
+    - name: main_attribute
+      dtype: string
+    - name: other_attributes
+      dtype: string
+    - name: bbox
+      dtype: string
+    - name: item_ID
+      dtype: string
+    - name: task
+      dtype: string
+    - name: difficulty
+      dtype: string
+---
+
+# LookBench: Fashion Image Retrieval Benchmark
+
+<div align="center">
+  <img src="assets/lookbench_banner.png" alt="LookBench Banner" width="800"/>
+</div>
+
+## 📋 Overview
+
+**LookBench** is a comprehensive benchmark for evaluating fashion image retrieval models. It provides a standardized evaluation framework with multiple difficulty levels and evaluation dimensions.
+
+### Key Features
+
+- 🎨 **Multi-task Evaluation**: 4 distinct retrieval tasks across different scenarios
+- 📊 **Three Evaluation Dimensions**: Coarse Recall, Fine Recall, and Graded NDCG
+- 🔊 **Noise Injection**: Large-scale distractor gallery for realistic evaluation
+- 🏷️ **Rich Metadata**: Category, attributes, bounding boxes, and difficulty levels
+
+## 📁 Dataset Structure
+
+```
+LookBench/
+├── v20251201/                    # Version: December 2024
+│   ├── aigen_streetlook/         # AI-generated street-style images
+│   │   ├── query.parquet
+│   │   └── gallery.parquet
+│   ├── aigen_studio/             # AI-generated studio images
+│   │   ├── query.parquet
+│   │   └── gallery.parquet
+│   ├── real_streetlook/          # Real street-style photos
+│   │   ├── query.parquet
+│   │   └── gallery.parquet
+│   ├── real_studio_flat/         # Real studio flat-lay images
+│   │   ├── query.parquet
+│   │   └── gallery.parquet
+│   └── noise/                    # Distractor images
+│       └── noise_*.parquet
+└── README.md
+```
+
+## 🔢 Data Schema
+
+| Column | Type | Description |
+|--------|------|-------------|
+| `image` | Image | Fashion product image |
+| `category` | string | Product category (e.g., "bag", "dress", "shoes") |
+| `main_attribute` | string | Primary visual attribute (e.g., "crocodile", "striped") |
+| `other_attributes` | string | Additional attributes (comma-separated) |
+| `bbox` | string | Bounding box coordinates [x1, y1, x2, y2] |
+| `item_ID` | string | Unique item identifier |
+| `task` | string | Task name |
+| `difficulty` | string | Difficulty level: "easy", "medium", "hard" |
+
+## 📊 Dataset Statistics
+
+| Task | Query | Gallery | Difficulty Distribution |
+|------|-------|---------|------------------------|
+| AIGen-StreetLook | 160 | 571 | Easy: 30%, Medium: 40%, Hard: 30% |
+| AIGen-Studio | 150 | 520 | Easy: 35%, Medium: 35%, Hard: 30% |
+| Real-StreetLook | 200 | 680 | Easy: 25%, Medium: 45%, Hard: 30% |
+| Real-StudioFlat | 180 | 610 | Easy: 30%, Medium: 40%, Hard: 30% |
+| **Noise** | - | ~57,000 | - |
+
+## 📏 Evaluation Metrics
+
+### 1. Coarse Recall@K
+Category-level matching. Returns 1 if any top-K result has the same category.
+
+### 2. Fine Recall@K
+Category + Main Attribute matching. Stricter criterion requiring both to match.
+
+### 3. Graded NDCG@K
+Full attribute scoring with graded relevance:
+- Base score (1.0) for category + main_attribute match
+- Bonus for each matching `other_attribute`
+
+## 🚀 Quick Start
+
+```python
+from datasets import load_dataset
+
+# Load a specific task
+dataset = load_dataset("your-org/LookBench", "aigen_streetlook")
+
+# Access query and gallery splits
+query_data = dataset["query"]
+gallery_data = dataset["gallery"]
+
+# View a sample
+print(query_data[0])
+```
+
+## 📈 Leaderboard
+
+| Rank | Model | Coarse@10 | Fine@10 | NDCG@10 | Avg |
+|------|-------|-----------|---------|---------|-----|
+| 1 | Model A | 85.2 | 72.1 | 68.5 | 75.3 |
+| 2 | Model B | 83.1 | 70.5 | 66.2 | 73.3 |
+| 3 | Model C | 81.5 | 68.2 | 64.1 | 71.3 |
+
+*Submit your results to be included in the leaderboard!*
+
+## 📖 Citation
+
+```bibtex
+@dataset{lookbench2024,
+  title={LookBench: A Comprehensive Benchmark for Fashion Image Retrieval},
+  author={Your Team},
+  year={2024},
+  publisher={Hugging Face},
+  url={https://huggingface.co/datasets/your-org/LookBench}
+}
+```
+
+## 📜 License
+
+This dataset is released under the [Apache 2.0 License](LICENSE).
+
+## 🔗 Links
+
+- [GitHub Repository](https://github.com/your-org/look-bench)
+- [Evaluation Code](https://github.com/your-org/look-bench)
+- [Paper (coming soon)]()
+
+---
+
+<div align="center">
+  <b>Version:</b> v20251201 | <b>Last Updated:</b> December 2024
+</div>
+

RELEASE_GUIDE.md

@@ -0,0 +1,182 @@
+# LookBench HuggingFace Release Guide
+
+This guide explains how to prepare and upload the LookBench dataset to Hugging Face.
+
+## Directory Structure
+
+```
+huggingface_release/
+├── README.md                    # Dataset card (displayed on HF)
+├── LookBench.py                 # Custom dataset loading script
+├── prepare_release.py           # Convert parquet to HF format
+├── upload_to_hf.py              # Upload to HuggingFace Hub
+├── create_sample_images.py      # Generate preview images
+├── assets/                      # Images for dataset card
+│   ├── lookbench_banner.png
+│   ├── evaluation_dimensions.png
+│   └── samples_*.png
+└── v20251201/                   # Versioned data
+    ├── aigen_streetlook/
+    │   ├── query.parquet
+    │   └── gallery.parquet
+    ├── aigen_studio/
+    │   ├── query.parquet
+    │   └── gallery.parquet
+    ├── real_streetlook/
+    │   ├── query.parquet
+    │   └── gallery.parquet
+    ├── real_studio_flat/
+    │   ├── query.parquet
+    │   └── gallery.parquet
+    ├── noise/
+    │   └── noise_*.parquet
+    └── version_info.json
+```
+
+## Release Steps
+
+### Step 1: Prepare the Data
+
+```bash
+cd huggingface_release
+
+# Convert existing parquet files to HF format
+python prepare_release.py \
+    --input_dir ../scripts/data/parquet_files_v5 \
+    --output_dir v20251201 \
+    --version v20251201
+```
+
+### Step 2: Create Assets
+
+```bash
+# Generate banner and sample images
+python create_sample_images.py \
+    --input_dir ../scripts/data/parquet_files_v5 \
+    --output_dir assets
+```
+
+### Step 3: Upload to HuggingFace
+
+```bash
+# Set your HuggingFace token
+export HF_TOKEN="your_token_here"
+
+# Upload using folder method (preserves structure)
+python upload_to_hf.py \
+    --repo_id your-org/LookBench \
+    --data_dir v20251201 \
+    --method api
+
+# Or upload using datasets library (creates proper configs)
+python upload_to_hf.py \
+    --repo_id your-org/LookBench \
+    --data_dir v20251201 \
+    --method datasets
+```
+
+## Versioning Strategy
+
+LookBench uses date-based versioning: `vYYYYMMDD`
+
+| Version | Date | Changes |
+|---------|------|---------|
+| v20251201 | Dec 2024 | Initial release |
+| v20250301 | Mar 2025 | (future) Updated gallery |
+
+To add a new version:
+1. Create a new versioned directory (e.g., `v20250301/`)
+2. Update `README.md` with new version info
+3. Add new config entries in the YAML frontmatter
+4. Keep old versions for reproducibility
+
+## Dataset Card (README.md)
+
+The README.md serves as the dataset card displayed on HuggingFace. Key sections:
+
+1. **YAML Frontmatter**: Metadata, configs, features
+2. **Overview**: Description and key features
+3. **Dataset Structure**: File organization
+4. **Data Schema**: Column descriptions
+5. **Statistics**: Sample counts and distributions
+6. **Evaluation Metrics**: How to evaluate
+7. **Leaderboard**: Model rankings
+8. **Citation**: BibTeX entry
+
+## Image Display on HuggingFace
+
+For images to display in the HuggingFace viewer:
+
+1. **Parquet format**: Store images as `{'bytes': image_bytes, 'path': None}`
+2. **Features definition**: Use `Image()` type in the schema
+3. **Dataset card images**: Reference with `![alt](assets/image.png)`
+
+### Parquet Image Format
+
+```python
+# Correct format for HF image display
+record = {
+    'image': {'bytes': image_bytes, 'path': None},
+    'category': 'bag',
+    ...
+}
+```
+
+### Dataset Card Images
+
+```markdown
+<!-- In README.md -->
+![Banner](assets/lookbench_banner.png)
+![Samples](assets/samples_aigen_streetlook.png)
+```
+
+## Loading the Dataset
+
+After upload, users can load the dataset:
+
+```python
+from datasets import load_dataset
+
+# Load a specific task
+ds = load_dataset("your-org/LookBench", "aigen_streetlook")
+
+# Access splits
+query = ds["query"]
+gallery = ds["gallery"]
+
+# View sample
+print(query[0])
+query[0]["image"]  # Returns PIL Image
+```
+
+## Updating the Leaderboard
+
+To update the leaderboard:
+
+1. Run evaluation on your model
+2. Submit results via GitHub issue or PR
+3. Update the leaderboard table in README.md
+
+## Troubleshooting
+
+### Images not displaying
+- Ensure images are in `{'bytes': ..., 'path': ...}` format
+- Check that Features specifies `Image()` type
+- Verify parquet was saved with correct schema
+
+### Upload fails
+- Check HF_TOKEN is valid and has write access
+- Ensure repo exists and you have permissions
+- Try uploading in smaller batches
+
+### Dataset won't load
+- Verify parquet files are valid: `pd.read_parquet(file)`
+- Check config names match directory structure
+- Ensure all required columns are present
+
+## Contact
+
+For issues or questions:
+- GitHub: https://github.com/your-org/look-bench
+- HuggingFace: https://huggingface.co/datasets/your-org/LookBench
+

v20251201/aigen_streetlook/gallery.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a216a5f8439f9df437c66cec0763ccf53ba295b874cc3ebd6a302ad430bf081
+size 2432918

v20251201/aigen_streetlook/query.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75eb347ad05f2bf83ebe114375c7d7e25c07062c4c6266d32d3f527437ca447e
+size 12946

v20251201/aigen_studio/gallery.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48fa3ee46be363813976874e92a4abb04041d85762ff456629c44ac37641c5e1
+size 4343641

v20251201/aigen_studio/query.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afa10a70baac8a035688b775f04777d4a33009ff24ed8d4ed3d16913770defc4
+size 600

v20251201/noise/noise_00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3db703367a4d08fd97899e1e709c59cc96c1825dc1c57926712dd1b2a9a62b2a
+size 152461313

v20251201/noise/noise_00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b0767ba2da7ea333ef9de2730823a9644b4af3d253bba02a8df9d5fe910f992
+size 153502246

v20251201/noise/noise_00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23a8688783b3f3d2914afe3da84607aa688602adf70cb0d0215d25eda2da3eb3
+size 138073565

v20251201/noise/noise_00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3e66e11e8f869303771335c957e520f60c1361bce56fb8f6b3158608c4246bd
+size 111873044

v20251201/noise/noise_00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9640fe9e30af8648f5d0d140b1dce148e39351b7a1656f64cf620cea6680aea1
+size 95035611

v20251201/noise/noise_00005.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c0ae8c1c348007bdc099f61f820fec2381e5cdb1e2a983f6cd6bf769de0869d
+size 181324854

v20251201/noise/noise_00006.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65edea341f0ed34f4687aff05f7b1332bc78c521c4f5fa358455fad25f5d552e
+size 204503584

v20251201/noise/noise_00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e7a65b44a23818b50ae764367a6aae2be4a72b6be0e48f6035be5ba8a8410ab
+size 204920425

v20251201/noise/noise_00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cf54619b4f6f9d9ffeed05c6f64a249505b613e9c13505dc500aafb95dedb4b
+size 163159466

v20251201/real_streetlook/gallery.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9379991a8a607771571e2c9b245ba9813747aa88371c88b6bcf3ee254662bbf5
+size 19647442

v20251201/real_streetlook/query.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:838db2199a217142425209329f352d9955c2c68417686ad942478d897b5c86af
+size 3466647

v20251201/real_studio_flat/gallery.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9918584999fb9de06d13ad79d720490a672a3a642f0784f8522fe166c6a604c
+size 30371576

v20251201/real_studio_flat/query.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6189792aa40991a096ac295ea07962be7dde4d2c2ef421276b405ffd5031ce9
+size 177085

v20251201/version_info.json

@@ -0,0 +1,11 @@
+{
+  "version": "v20251201",
+  "created_at": "2025-12-10T15:28:37.912666",
+  "tasks": [
+    "aigen_streetlook",
+    "aigen_studio",
+    "real_streetlook",
+    "real_studio_flat"
+  ],
+  "includes_noise": true
+}