这个地方代码有错吧，image是个1*3*h*w的 tensor张量shape取值是错的 ，而且image不需要再.unsqueeze(0) image_features = [] for image in images: image = image.unsqueeze(0) P, C, H, W = image.shape P, C, H, W = image.shape ValueError: too many values to unpack (expected 4)
这个地方代码有错吧，image是个13h*w的 tensor张量shape取值是错的 ，而且image不需要再.unsqueeze(0)
image_features = []
for image in images:
image = image.unsqueeze(0)
P, C, H, W = image.shape
P, C, H, W = image.shape
ValueError: too many values to unpack (expected 4)