Closed Dhyey-Savaliya closed 3 years ago
How can I change input shape to predict class for one dimensional images??
transform N,C,H,W to N,C, HW. And then change conv2d, bn2d to conv1d and bn1d. And you cannot load the imagenet pretrain when using 1d version of res2net.
How can I change input shape to predict class for one dimensional images??