search

PaddlePaddle / PaddleDetection

Object Detection toolkit based on PaddlePaddle. It supports object detection, instance segmentation, multiple object tracking and real-time multi-person keypoint detection.
Apache License 2.0
12.65k stars 2.87k forks source link

[Other General Issues]关于ppyoloe 验证步骤输入图像target_size大小限制的问题 #5498

Closed Tongfengyu closed 2 years ago

Tongfengyu commented 2 years ago

验证读入图片resize到256会报错如下： 屏幕截图 2022-03-29 152539 调到默认640，不会报错 1 Q:1.不知道为什么，是输入图片不能太小吗？2.Y=[8400,2],第一维度的8400是什么意思，在哪定义的？

lzh111222334 commented 2 years ago

+1，推理模型导出的时候也遇到这个问题

jerrywgz commented 2 years ago

目前ppyoloe如果要设置其他尺寸评估，除了修改EvalReader之外，还需要修改https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.4/configs/ppyoloe/base/ppyoloe_crn.yml#L31 中的eval_input_size为对应尺寸，或者不设置eval_input_size 这个我们后面再优化下

Tongfengyu commented 2 years ago

谢谢

sdreamforchen commented 2 years ago

您好。我在转ONNX模型的时候，提示了警告：Due to the operator:multiclass_nms3, the converted ONNX model will only supports input[batch_size] == 1； 我得到的onnx，我查看输入的维数为[-1,3,640,640]，下游环境对这个-1不能识别，比较悲催。请问可以在哪设置一下吗

paddle-bot-old[bot] commented 2 years ago

Since this issue has not been updated for more than three months, it will be closed, if it is not solved or there is a follow-up one, please reopen it at any time and we will continue to follow up. It is recommended to pull and try the latest code first. 由于该问题超过三个月未更新，将会被关闭，若问题未解决或有后续问题，请随时重新打开（建议先拉取最新代码进行尝试），我们会继续跟进。