This repo is NVDLA for low precision exploration. I will try to add more test. (I am preparing my master's thesis, so the project is temporarily suspended)
- Vistualization model
- Yolo Object Detection (Caffe)
- Cifar10 Quick every layer vistualization (Caffe)
- Caffe result and NVDLA result compare
- Cifar10 Quick first convolution feature maps (Random sample 42 test/train images)
- Cifar10 Quick prediction results (Random sample 42 test/train images)
- Others
- Read image mean subtraction by pixel vistaulization