Pipeline based on Open-MMLAB MM-Detection project - https://github.com/open-mmlab/mmdetection
- faster_rcnn_fpn50
- faster_rcnn_fpn101
- faster_rcnn_x101_32x4d_fpn
- faster_rcnn_x101_64x4d_fpn
- cascade_rcnn_fpn50
- cascade_rcnn_fpn101
- cascade_rcnn_x101_32x4d_fpn
- cascade_rcnn_x101_64x4d_fpn
- retinanet_r50_fpn
- retinanet_r101_fpn
- retinanet_x101_32x4d_fpn
- retinanet_x101_64x4d_fpn
- retinanet_ghm_r50_fpn
- retinanet_ghm_r101_fpn
- retinanet_ghm_x101_32x4d_fpn
- retinanet_ghm_x101_64x4d_fpn
- dh_faster_rcnn_fpn50
- libra_faster_rcnn_fpn50
- libra_faster_rcnn_fpn101
- libra_faster_rcnn_x101_64x4d_fpn
- libra_retinanet_r50_fpn
- ga_faster_rcnn_x101_32x4d_fpn
- ga_faster_rcnn_x101_64x4d_fpn
- ga_retinanet_x101_32x4d_fpn
- ga_retinanet_x101_64x4d_fpn
- fovea_r50_fpn_4x4
- fovea_r101_fpn_4x4
- fovea_align_r50_fpn_gn-head_mstrain_640-800_4x4
- fovea_align_r101_fpn_gn-head_mstrain_640-800_4x4
- free_anchor_retinanet_r50_fpn
- free_anchor_retinanet_r101_fpn
- free_anchor_retinanet_x101_32x4d_fpn
- atss_r50_fpn
- pafpn_faster_rcnn_r50
- faster_rcnn_r50_fpn_mdpool
- faster_rcnn_r50_fpn_dpool
Supports
- Python 3.6
- Cuda 9.0, 10.0 (Other cuda version support is experimental)
cd installation
chmod +x install.sh && ./install.sh
- Load Dataset
gtf.Train_Dataset(img_dir, annofile, class_file);
gtf.Dataset_Params(batch_size=2, num_workers=2);
- Load Model
gtf.Model_Params(model_name="faster_rcnn_x101_64x4d_fpn");
- Set Hyper Parameters
gtf.Hyper_Params(lr=0.02, momentum=0.9, weight_decay=0.0001);
gtf.Training_Params(num_epochs=2, val_interval=1);
- Train
gtf.Train();
- Add support for Coco-Type Annotated Datasets
- Add support for VOC-Type Annotated Dataset
- Test on Kaggle and Colab
- Add validation feature & data pipeline
- Add Optimizer selection feature
- Enable Learning-Rate Scheduler Support
- Enable Layer Freezing
- Set Verbosity Levels
- Add Project management and version control support (Similar to Monk Classification)
- Add Graph Visualization Support
- Enable batch proessing at inference
- Add feature for top-k output visualization
- Add Multi-GPU training
- Auto correct missing or corrupt images - Currently skips them
- Add Experimental Data Analysis Feature