SqueezeNet-SSD icon indicating copy to clipboard operation
SqueezeNet-SSD copied to clipboard

SSD detection network of SqueezeNet, with pretrained weights on VOC0712 and mAP=0.643

Results 9 SqueezeNet-SSD issues
Sort by recently updated
recently updated
newest added

@chuanqi305 Hi, I use all same file you provide for trianing, but just get 0.4 mAP on VOC0712, can you give some advice about how you train squeezenet ssd 300*300....

Hi, I'm looking to train my own data using 2 classes, I was wondering which parameters do I need to change in the train, test and deploy prototxt file. I've...

command: python merge_bn.py --model squeezenet_ssd_voc_deploy.prototxt --weights squeezenet_iter_74000.caffemodel result: Unknown bottom blob 'fire5/normal' (layer 'fire5_mbox_loc', bottom index 0)

Hi, @chuanqi305 In README.md, you say > with pretrained weights on VOC0712 and mAP=0.643 I found that this sentence is similar to the one in mobilenet-ssd, In mobilenet-ssd's README.md, it...

Hi, @chuanqi305 please tell me How to make squeezenet_iter_74000.caffemodel. I used https://github.com/chuanqi305/SqueezeNet-SSD/blob/master/solver_train.prototxt and got squeezenet_iter_220000.caffemodel. but test result with solver_test.prototxt is about 42.2 mAP. thx..

where do you get the pretrained weights on VOC0712? Could you train the VOC0712 using the SqueezeNet?

I am confused about solver_train.prototxt and solver_test.prototxt. - different base_lr set, lr=0.0001 in solver_train.prototxt and lr=0.001 in solver_test.prototxt - different gamma set, gamma=0.5 in solver_train.prototxt and gamma=0.1 in solver_test.prototxt -...

Hi Chuanqi, I found the min_size and max_size value in squeezenet_ssd_voc_train.prototxt for prior boxes are different with the values in SSD. Do you make some difference when computing the values?...

Hi: can you provide an example for testing the model? thanks