Custom data training using VGG16 deploy.prototxt and .caffemodel

chuanqi305 / MobileNet-SSD

Caffe implementation of Google MobileNet SSD detection network, with pretrained weights on VOC0712 and mAP=0.727.

MIT License

2k stars 1.18k forks source link

Open Ram-Godavarthi opened 4 years ago

Ram-Godavarthi commented 4 years ago

Hello,

Anybody has successfully done training using Caffe VGG models on custom dataset?? I am finding it difficult to do it.

Could you please guide me through the repo or steps to do it .

I have images with 37 classes to be detected.

I would like to do it and run inference in TensorRT sampleSSD where it supports Vgg16 Caffe model.

Any help would be appreciated