Yolo-Fastest
Yolo-Fastest copied to clipboard
:zap: Based on yolo's ultra-lightweight universal target detection algorithm, the calculation amount is only 250mflops, the ncnn model size is only 666kb, the Raspberry Pi 3b can run up to 15fps+, and...
目前嘗試不使用預設的detector 權重開始,訓練自建的資料集。結果會遇到nan,特別是在第二段的v3 loss輸出,會觀察到某一段時間後,loss值直接飆上去,然後就爆掉了。連帶影響到第一段v3 loss變化。 請問各位有遇過類似的狀況嘛? 訓練機器偏小,弱弱的一張1050。跑ubuntu 20.04。使用darknet進行訓練yolo-fastest。自訂的訓練集,有50類,每類約有3000個框框。
Hi @dog-qiuqiu thanks for your great project. I'm just wondering, would it be possible to share your pedestrian's data set used in your project?
当训练小分辨率的模型时,在压缩到小分辨率的图像上,原本就比较小的人脸会变得极小,可能只有4、5个像素大小,请问怎么对待这些人脸呢? 1、如果设定比较小的anchor 4~10个像素的,这么小,感觉比较难训吧? 2、如果不想训练小于10个像素的,那应该是怎么对待呢,忽略(loss=0)还是当作负样本?
当训练小分辨率的模型时,在压缩到小分辨率的图像上,原本就比较小的人脸会变得极小,可能只有4、5个像素大小,请问怎么对待这些人脸呢? 1、如果设定比较小的anchor 4~10个像素的,这么小,感觉比较难训吧? 2、如果不想训练小于10个像素的,那应该是怎么对待呢,忽略(loss=0)还是当作负样本?
大佬您好,我想把yolo-fastest的.weights模型转换为.h5部署,使用yolov3.weights的转换文件修改网络文件之后转换报错。请问该如何解决? File "convert.py", line 263, in _main(parser.parse_args()) File "convert.py", line 143, in _main buffer=weights_file.read(weights_size*4)) TypeError: buffer is too small for requested array
I want to know how to convert the model weight to onnx model,can you help me?thanks
Where can I download yolo-fastest.conv.109 used for training? I can't find it.
Could you release the code of training pedestrian detection model with yolo-fastest. Thanks a lot
As the README.md said: Yolo-Fastest-1.1-xl has 0.925M Params, but I don't get it how to calculate it? If the model has been build on keras, I can use API model.summary()...