SeqFormer
SeqFormer copied to clipboard
SeqFormer: Sequential Transformer for Video Instance Segmentation (ECCV 2022 Oral)
This PR removes the `self.output_proj_box` from `MSDeformAttn` if the mode is `decode` which allows to run `torch.nn.parallel.DistributedDataParallel` without `find_unused_parameters=True`. In theory, this should improve training time as the torch backend...
The T=1 pretrained model files do not contain the `time_attention_weights` weights.
Following the README.md, the `ytvis` dataset folder will be in the root directory of this repository. Excecuting the `inference.py` script in the same directory will cause an error as it...
Congrats for the awesome work. I am trying to reproduce the results for resnet-50 backbone. I tried following , 1. Train Seqformer on coco dataset (with num_frames=1) for 24 epochs...
Hi, thanks for your good work. I want to know the performance only using the original DETR instead of the improved Deformable DETR for a fair comparison with IFC paper.
I try to run python reference.py,but I only got a json file.
How can we test on inference especially I'm getting errors on size? I have downloaded the pre-trained weights for r50 from readme.md and put backbone as resnet50. I'm getting size...