背单词
背单词
thank u,i will try again :>
我也遇到了相同的问题,请问您解决了吗?
thank u,i will try again :)
WoW,thank u very much,it work~
ffmpeg的问题已经解决,现在遇到这个问题 data = np.array(imageio.mimread(path, memtest=False), dtype=object)[..., :3] ValueError: could not broadcast input array from shape (280,260,3) into shape (280,260) 请问这个是要改为灰度图嘛?
给出的错误是数据不规范,我又试着遍历发现第一帧的形状是 (280, 260, 3),但是后面的形状都是 (280, 260, 4)就很奇怪 
问题都已经解决,对于论文中源代码 data = np.array(imageio.mimread(path, memtest=False), dtype=object)[..., :3] 我修改为: data = np.array(imageio.mimread(path, memtest=False,format='GIF-PIL'), dtype=object)[..., :3] 就能够正确的读取GIF图片的通道数(280,260,1)
 还有一点很奇怪,最后的输出结果是这样的
感谢你的回复,我会尝试继续测试,再次感谢:) Thank you for your reply. I will try to continue testing. Thank you again: : )
在阅读代码时我发现论文使用了22个人体关节点,请问为什么不是原数据的25个呢? When reading the code, I found that the paper used 22 human joint points. May I ask why it is not the original data of 25?