Deep network structure based projects, including source code, analyses and report. Most of them contain PyTorch and Tensorflow two versions.
- Have included the deep network for playing with cifar10/100 data to achieve object classification.
- Uploaded simple FasterRCNN network. (Detect cifar10 with the cifar100 as background).
- Completed GAN series, including DCGAN, WGAN and WGAN-GP for dataset CUFS and celebA.
- Uploaded 3D object detection network for 3D MNIST dataset.
- Uploaded AE, VAE network for dataset CUFS and celebA.
- Completed Image to Image network for dataset CUFS and Facades.
- Completed Unpaired image to image using CycleGAN.