Yang Xi
Yang Xi
Hello, I've seen you mention the LPCB article multiple times. In this article, it is mentioned that using standard cross-entropy to train HED produces a result similar to a relief...
checkpoints文件夹下
能把你预测结果发给我看看吗
基于卷积神经网络得到的边缘一般都比较粗，需要用非极大值抑制才能得到更细的边缘，如果你要更细的边缘的话你可以看看用用扩散模型来进行边缘检测那篇文章，他得到的边缘很细
https://paperswithcode.com/paper/diffusionedge-diffusion-probabilistic-model
> 在mamba-1p1p1/mamba_ssm/ops/selective_scan_interface.py文件里，`conv1d_out = causal_conv1d_cuda.causal_conv1d_fwd(x, conv1d_weight, conv1d_bias,None, True)` 改为 `conv1d_out = causal_conv1d_cuda.causal_conv1d_fwd(x, conv1d_weight, conv1d_bias, True)`，还有这里也修改 ``` dx, dconv1d_weight, dconv1d_bias = causal_conv1d_cuda.causal_conv1d_bwd( # x, conv1d_weight, conv1d_bias, dconv1d_out, None, dx, True x, conv1d_weight,...
如果是白底图片，评估代码中的is_white_edge设为false，如果是黑底图片，is_white_edge设为ture