How can we train on our own custom dataset?

[zzzxxxttt]

Just modify the datasets/coco.py (or create a new dataset file) and implement the image reading/preprocessing part in __getitem__ function, the output of __getitem__ should follow the same pattern:

        return {'image': img,
            'hmap': hmap, 'w_h_': w_h_, 'regs': regs, 'inds': inds, 'ind_masks': ind_masks,
            'c': center, 's': scale, 'img_id': img_id}

[davidhuangal]

Thank you.