Giters

yan-hao-tian / ConTNet

This repo contains the code of "ConTNet: Why not use convolution and transformer at the same time?"

Home Page:https://arxiv.org/abs/2104.13497

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Questions about "RuntimeError: The size of tensor a (238) must match the size of tensor b (243) at non-singleton dimension 3"

Simon-Stma opened this issue · comments

commented

In the summation of residual and identity, there is a dimensional mismatch, and after debugging, it is found that the dimensions will be different after transformer, how to solve this problem?

commented

To add to the above question, it is well known that the fasterrcnn task is not required to specify the resolution of the input images (he has default intervals), but after referring to the code in the article, I think of him more as a paradigm for a classification task based on 224x224 input images, so I would like to know how you applied ConTNet to the fasterrcnn detection task? How did you handle it for fasterrcnn input images with different resolutions?