libfacedetection.train
libfacedetection.train copied to clipboard
ShiqiYu
The training program for libfacedetection for face detection and 5-landmark detection.
I encountered an issue while attempting to train the Yunet model. Following the repository's instructions, I successfully cloned the repository. However, I faced a roadblock when I tried to proceed,...
hi, 我check 代码时发现, 你们的测试图未做归一化, 一般是为了抑制环境,光线等干扰因素. 为什么你们不用做归一化处理? 请帮忙解答. 多谢! dict( type='Normalize', mean=[0., 0., 0.], std=[1., 1., 1.], to_rgb=False),
I retrained the model with 16 batches. Then I converted to ONNX. **ONNX Sample** ort_session = ort.InferenceSession("yunet_16.onnx", providers=['CUDAExecutionProvider']) input_name = ort_session.get_inputs()[0].name for i_path in tqdm(os.listdir(images_path)): for i in os.listdir(os.path.join(images_path, i_path)):...
The library right now works great for detecting faces, but the landmarks are clearly poorly trained. They are by no means accurate and are often not even close to being...
``` Pytorch: 1.13.1 cuda: 11.6 mmcv:1.6.0 ``` RuntimeError: Only tuples, lists and Variables are supported as JIT inputs/outputs. Dictionaries and strings are also accepted, but their usage is not recommended....
Hi, I'm following the README.md instructions for installing everything on my computer, but I'm facing some kind of error when I'm trying to execute the training process. I've installed pytorch...
Did you try Focal loss, how about its performance ?
https://github.com/ShiqiYu/libfacedetection.train/raw/master/tasks/task1/weights/yunet_final.pth. this window not found and how can i find the yunet_final.pth weight file.
余老师您好，我在下载了您的开源项目之后想要尝试把onnx模型部署到VUE上并做一个demo的展示。目前我使用的Node.js是1.16.1，onnxruntime是最新的1.12.1.在这个过程中我发现，使用项目自带的onnx会遇到两个问题： 1.需要把INT64转换成INT32.这个比较好处理。我在opset_version等参数不变的情况下，把模型的params和Nodes都换成了精度是INT32的。 2.然后，我就发现，您提供的ONNX似乎在JS上仍然不是很兼容。具体错误如下图：  我还尝试在您的代码上把上采样的方法改写，不进行任何操作而直接返回一个符合上采样尺寸大小的torch.ones()。但前端还是会出现Shape的错误。这个有什么好方法或者建议去解决吗？ 
Hi. I saw the new version of libfacedetection.train, which is switched to the mmdet framework. But both the yunet_n.py and yunet_s.py are little bit different from the original model structure....
