Ashutosh Mishra

Loss value & LR constant

10

Hi @zylo117 I am now performing finetuning after my previous issue got fixed. Thanks for your help! I have a total of 56K images of road scenes trying to detect...

ONNX/Caffe2 model conversion issue

1

Hi, Thanks for the repo containing all the models! While converting pretrained mobilnet v1 SSD model to caffe2 and onnx, there is no predictor layer attached at the end because...

Hi, Thanks for this wonderful repo! I would like to ask you whether you have trained Cityscapes images on full resolution images using DeeplabV3 + Mobilenet architecture model you have...

Pytorch->ONNX->TF-> SNPE Error for batchnorm variance

3

I converted Mobilenet V1 SSD Pytorch model to ONNX format(https://github.com/qfgaohao/pytorch-ssd) and from there to protobuf format using ONNX TF (TF==1.15). ONNX version == 1.3 and SNPE == 1.36 Now when...

Hi, Thanks for providing the code !! Can you please provide the training code so as to replicate the results for ESNet !! Thanks !!

Tracking vs Identification

1

Hi, Thanks for providing this wonderful repository! My doubt is that in real-world setting, where you have multiple persons entering the camera feed and you need to ID them for...

cudnn.hpp(126): Error: argument of type "int" is incompatible with parameter of type "cudnnNanPropagation_t"

1

I am trying to compile the Caffe submodule version but it gives the above mentioned error when building with cmake. I am using CUDA 8.0 and cuDNN 5.1.5 Does this...

I am running the code on cityscapes dataset which has image, labels but not weights. I am passing a list of ones as argument to weight list in read_file function....

Overlapping instances in fine-tuning SAM

2

Hello @bnsreenu Thanks for providing the repo for fine-tuning the SAM model on the custom dataset. My question is that in case where the instances of the object are overlapping...

I have some .mat files which I need to convert into lmdb format. Could you please tell how can I do it using this wrapper?