| | --- |
| | license: apache-2.0 |
| | tags: |
| | - image-classification |
| | - object-detection |
| | - instance-segmentation |
| | - medical-image-segmentation |
| | - graph-neural-network |
| | - mobile |
| | - vision-gnn |
| | - pytorch |
| | datasets: |
| | - imagenet-1k |
| | - coco |
| | - kvasir-seg |
| | - dsb2018 |
| | metrics: |
| | - top_1_accuracy |
| | - mean_average_precision |
| | - dice_score |
| | - hausdorff_distance |
| | library_name: pytorch |
| | pipeline_tag: image-classification |
| | --- |
| | |
| | # GeoViG: Geometry-Aware Graph Reasoning for Mobile Vision Tasks |
| |
|
| | ## Model Variants |
| |
|
| | ### 🖼️ Image Classification — ImageNet-1K |
| |
|
| | | Model | Params (M) | MACs (G) | Top-1 Acc (%) | Checkpoint (pth) | |
| | |:---|:---:|:---:|:---:|:---:| |
| | | GeoViG-Ti | 3.5 | 0.9 | 75.2 | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/pth/geovig_ti_5e4_8G_300_75_22/checkpoint.pth) | |
| | | GeoViG-S | 5.0 | 1.2 | 77.5 | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/pth/geovig_s_5e4_8G_300_77_48/checkpoint.pth) | |
| | | GeoViG-M | 10.3 | 2.2 | 80.7 | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/pth/geovig_m_5e4_8G_300_80_70/checkpoint.pth) | |
| | | GeoViG-B | 19.7 | 4.5 | 82.4 | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/pth/geovig_b_5e4_8G_300_82_38/checkpoint.pth) | |
| |
|
| | Trained for 300 epochs on 8× NVIDIA A100 GPUs, batch size 1024, AdamW optimizer (lr=5e-4, weight decay=0.05), cosine schedule with 5-epoch warmup. Augmentations: RandAugment, Mixup (p=0.8), CutMix (p=1.0). |
| |
|
| | For CoreML checkpoint, please check: https://huggingface.co/OmarAlasqa/GeoViG/tree/main/CoreML |
| | For IPA chechpoints, please check: https://huggingface.co/OmarAlasqa/GeoViG/tree/main/IPA |
| | --- |
| |
|
| | ### 📦 Object Detection & Instance Segmentation — MS COCO 2017 |
| |
|
| | Backbone used with **Mask R-CNN**, 1× schedule (12 epochs), pretrained on ImageNet-1K. |
| |
|
| | | Backbone | Params (M) | Box AP | Box AP₅₀ | Box AP₇₅ | Mask AP | Mask AP₅₀ | Mask AP₇₅ | Checkpoint | |
| | |:---|:---:|:---:|:---:|:---:|:---:|:---:|:---:|:---:| |
| | | GeoViG-M | 10.3 | 40.7 | 62.4 | 44.1 | 37.7 | 59.6 | 40.5 | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/coco_det_seg_pth/geovig_m_det_seg/epoch_12.pth) | |
| | | GeoViG-B | 19.7 | 42.5 | 64.0 | 46.8 | 38.9 | 61.2 | 41.7 | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/coco_det_seg_pth/geovig_b_det_seg/epoch_12.pth) | |
| |
|
| | --- |
| |
|
| | ### 🏥 Medical Image Segmentation |
| |
|
| | **Kvasir-SEG — Polyp Segmentation** |
| |
|
| | | Model | Params (M) | mAP | Dice ↑ | IoU ↑ | Hausdorff Dist ↓ | Checkpoint | |
| | |:---|:---:|:---:|:---:|:---:|:---:|:---:| |
| | | GeoViG-M | 29.57 | 0.990 | **0.945** | **0.909** | **12.94** | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/medical/kvasir_geovig_m/checkpoint.pth) | |
| |
|
| | **Data Science Bowl 2018 — Nuclei Segmentation** |
| |
|
| | | Model | Params (M) | mAP | Dice ↑ | IoU ↑ | Hausdorff Dist ↓ | Checkpoint | |
| | |:---|:---:|:---:|:---:|:---:|:---:|:---:| |
| | | GeoViG-M | 29.57 | 0.859 | **0.908** | **0.839** | **5.19** | [Download](https://huggingface.co/OmarAlasqa/GeoViG/blob/main/medical/dsb_geovig_m/checkpoint.pth) | |
| |
|
| | --- |
| |
|
| | ## 🚀 Usage |
| |
|
| | Please check the github repo: https://github.com/OmarAlsaqa/GeoViG |
| |
|
| | --- |
| |
|
| | ## Citation |
| |
|
| | If you use GeoViG in your research, please cite: |
| |
|
| | ```bibtex |
| | @article{alsaqa2026geovig, |
| | title = {GeoViG: Geometry-Aware Graph Reasoning for Mobile Vision Tasks in Natural and Medical Images}, |
| | author = {Alsaqa, Omar and Mohammed, Emad and Aleem, Saiqa}, |
| | journal = {Under Review at IEEE EMBC}, |
| | year = {2026} |
| | } |
| | ``` |
| |
|
| | --- |
| |
|
| | ## Acknowledgements |
| |
|
| | This work builds upon [MobileViG](https://github.com/SLDGroup/MobileViG) and uses the [MMDetection](https://github.com/open-mmlab/mmdetection) framework for detection and segmentation experiments. Training was performed on the Compute Canada A100 cluster. |
| |
|
| | --- |
| |
|
| | ## License |
| |
|
| | This project is released under the [Apache 2.0 License](LICENSE). |
| |
|
