CRAFT-pytorch
CRAFT-pytorch copied to clipboard
clovaai
Official implementation of Character Region Awareness for Text Detection (CRAFT)
I implemented the inference code for CRAFT using C++ and used TensorRT as the inference backend. The code is available at: [https://github.com/yinnhao/CRAFTcpp](https://github.com/yinnhao/CRAFTcpp). Anyone interested can check and use it.
https://demo.ocr.clova.ai/ The live demo mentioned in the readme is down
Had this issue. in detect_text -> def get_prediction : in this line: `polys = craft_utils.adjustResultCoordinates(polys, ratio_w, ratio_h)` ``` def adjustResultCoordinates(polys, ratio_w, ratio_h, ratio_net=2): if len(polys) > 0: polys = np.array(polys)...
edit vgg16_bn.py. forward() *because torch jit does not supply "named tuple" ```python def forward(self, X): h = self.slice1(X) h_relu2_2 = h h = self.slice2(h) h_relu3_2 = h h = self.slice3(h)...
This fix enables garbage collection to appropriately work when https://github.com/clovaai/CRAFT-pytorch/blob/e332dd8b718e291f51b66ff8f9ef2c98ee4474c8/test.py#L69 returns, by deleting `x` we moved to the GPU after we move the forward pass results back to the CPU....
### Description This pull request fixes issue #194 by adding `dtype=object` when creating the `polys` array in the `adjustResultCoordinates()` function. This change allows the creation of arrays with incompatible types,...
