search

Epiphqny / VisTR

[CVPR2021 Oral] End-to-End Video Instance Segmentation with Transformers
https://arxiv.org/abs/2011.14503
Apache License 2.0
739 stars 95 forks source link

pretrained Detr weights #58

Closed helq2612 closed 3 years ago

helq2612 commented 3 years ago

Hi,

Thank you for your great work! Do you pre-train the Detr weights by yourself? Because I find the official Detr hidden_dim is 256, not 384. If so, do you train it with the object detection task, or the panoptic task?

Epiphqny commented 3 years ago

Hi @helq2612 I trained with the instance segmentation task, with the hidden_dim being 384.

helq2612 commented 3 years ago

Thank you very much!