luzhenyu
luzhenyu
Baofengzan仓库里是换成了最近邻上采样，方便tensorrt部署，但双线性上采样也是支持的啊，两者精度会相差多少？有小伙伴试过吗？
https://github.com/xiaoyufenfei/Efficient-Segmentation-Networks/blob/0f0c32e7af3463d381cb184a158ff60e16f7fb9a/utils/metric/metric.py#L43-L56 it's reverse. ` # 理解混淆矩阵生成的代码的关键，行为真实值，列为预测值 # Pii为预测正确的数量，Pij=FN, Pji=FP # 每一列之和表示被预测为该类别的样本数量 = TP+FP, precision = TP/(TP+FP), 所有被预测为正类中真正正类的比例 # 每一行之和表示该类别的真实样本数量 = TP+FN, recall = TP/TP+FN , 所有正类中，被找出的正类的比例 def recall(self): recall =...
[https://github.com/Reagan1311/DABNet/blob/b8d62fe7f14ae4909a9e9aad1dd6e0ade98431cd/utils/metric.py#L43-L56](url) it's reverse. ` # 理解混淆矩阵生成的代码的关键，行为真实值，列为预测值 # Pii为预测正确的数量，Pij=FN, Pji=FP # 每一列之和表示被预测为该类别的样本数量 = TP+FP, precision = TP/(TP+FP), 所有被预测为正类中真正正类的比例 # 每一行之和表示该类别的真实样本数量 = TP+FN, recall = TP/TP+FN , 所有正类中，被找出的正类的比例 def recall(self): recall =...
在Ultra-Light-Fast-Generic-Face-Detector-1MB中推理没有用到prior_box，对照ssd，你的代码是是用到了prio_box，这样在推理时会增加计算量，这样做是为了更好的训练结果吗？
I test the 78th model(training batchsize=1,the validation's psnr is 29.58) on Foliage from Vid4, the psnr between sr and hr is around 20. The result in the frvsr paper is...
I wonder what the front 2 frames of the first frame are? In the DUF, I see it's np.pad 0.
I see def conv_init(m): in the wide_resnet.py of models, but it just be defined, and it's not applyed. Is it should be used as net.apply(conv_init)?
### 问题确认 Search before asking - [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer. ### 请提出你的问题 Please ask your question 我用关键词fasterrcnn检索了open或closed的问题，但是没能找到我想要的答案。 我是想问，就是简单的按照deploy/cpp/docs/windows_vs2019_build.md里面的操作，就可以了吗？ 看了include里的代码，都是头文件，是不是官方不开源这部分代码？
ERROR
[INFO] Init cuBLASLt cosine similarity calculator... terminate called after throwing an instance of 'nlohmann::detail::type_error' what(): [json.exception.type_error.305] cannot use operator[] with a numeric argument with object Aborted (core dumped) 运行环境：Jetson Xavier...