lewisnjue
FoodSeg103 is a new food image dataset containing 7,118 images. Images are annotated with 104 ingredient classes and each image has an average of 6 ingredient labels and pixel-wise masks. It's provided as a large-scale benchmark for food image segmentation.
Major Challenges:
Image source: https://arxiv.org/pdf/2105.05409v1.pdf
Variants: FoodSeg103
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Semantic Segmentation | FoodSAM | FoodSAM: Any Food Segmentation | 2023-08-11 |
| Semantic Segmentation | CCNet (ReLeM-ResNet-50) | A Large-Scale Benchmark for Food … | 2021-05-12 |
| Semantic Segmentation | SeTR-Naive (ReLeM-ViT-16/B) | A Large-Scale Benchmark for Food … | 2021-05-12 |
| Semantic Segmentation | Swin-Transformer (Swin-Small) | Swin Transformer: Hierarchical Vision Transformer … | 2021-03-25 |
| Semantic Segmentation | SeTR-Naive (ViT-16/B) | Rethinking Semantic Segmentation from a … | 2020-12-31 |
| Semantic Segmentation | SeTR-MLA (ViT-16/B) | Rethinking Semantic Segmentation from a … | 2020-12-31 |
| Semantic Segmentation | CCNet (ResNet-50) | CCNet: Criss-Cross Attention for Semantic … | 2018-11-28 |
Recent papers with results on this dataset: