search

alanlukezic / d3s

D3S - Discriminative Single Shot Segmentation Tracker (CVPR 2020)
265 stars 57 forks source link

About object boundingbox label. #1

Closed deepbrainme closed 4 years ago

deepbrainme commented 4 years ago

I have clone your code and vos dataset(version2018), but I came across a run error. It show me that my dataset don't have a file in named 'groundtruth-obj_id.txt', in each seqence in youtube-vos dataset. My dataset have some wrong or miss some files ?

def _read_anno(self, seq_path, obj_id): anno_file = os.path.join(seq_path, "groundtruth-%s.txt" % obj_id) gt = pandas.read_csv(anno_file, delimiter=',', header=None, dtype=np.float32, na_filter=False, low_memory=False).values return torch.tensor(gt)

alanlukezic commented 4 years ago

Thank you for pointing this out, I forgot to include some files. The VOS dataset does not have bounding boxes, so I generated them from segmentation masks. You can download them on the link below and copy them in the dataset. Then it should work. Bounding boxes download link: http://data.vicos.si/alanl/d3s/rectangles.zip

deepbrainme commented 4 years ago

Thanks for your reply ,I copy recatngles.zip to train/Annotations/ ,and unzip it. Now it works.