Li

NNCLR：queue_size

2

Hi, In paper " With a Little Help From My Fridens：Nearest-Neighbor Contrastive Learning of Visual Representations", the queue size of the support is kept large enough so as to approximate...

Training_Loss

1

您好，请问您readme.md中的demo展示是使用哪种backbone得到的结果呢？还有在train过程中仅使用了wing_loss嘛？

Hi, When I train the video modality, I can not get a good result, there may some errors be in the prepared the video labels?

Hi, thanks for your excellent work. When I rerun this code, I get some problem: the training loss is descending and the ACC is improving, but the test ACC is...

loss项

1

`def calculate_loss(predict_keypoints, label_keypoints): landmark_label = label_keypoints[:, 0:136] pose_label = label_keypoints[:, 136:139] leye_cls_label = label_keypoints[:, 139] reye_cls_label = label_keypoints[:, 140] mouth_cls_label = label_keypoints[:, 141] big_mouth_cls_label = label_keypoints[:, 142] landmark_predict = predict_keypoints[:,...

实际检测准确率

2

您好，请问这个实际检测的准确率很差，是我自己的问题嘛。大家有在端侧或虚拟机上尝试嘛？ 

Hi, thanks for your work. In the 'shufflenetv2.py', I see that ' Input size needs to be divisible by 32', such as 96 ... so we do this is only...

随着训练，显存占用增大？

7

哈喽，老师~ 请问您有遇到过随着训练epoch的增加，显存逐渐增大的问题嘛？ 我更换了backbone后，会不断出现这样的情况~~

PP-PicoDet推理延迟

3

### 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  您好，请问我在虚拟机上使用官方给的ckpt与ncnn对应的demo代码进行推理延迟测试，结果怎么跟论文中给出结果差这么多~是哪里有问题嘛？我自己也确认了代码好像没啥问题。。

端侧推理

7

您好，请问有做arm端的推理延迟对比嘛？ Jetson AGX Xavier这个芯片算力比一般普通芯片高太多了，如高通、联发科的一些arm端芯片