Zheng Dong

representation learning (1) 3d reconstruction (1) multi-task learning (1) offline reinforcement learning (1) feature extraction (1) latent dirichlet allocation (1) topic modeling (1) 3d human reconstruction (1) deep learning (1) image processing (1) convolutional neural network (1) perceptual quality (1) model generalization (1) value network (1) spatiotemporal forecasting (1) masked autoencoder (1) multi-scale feature (1) self-supervised pre-training (1) depth denoising (1) recurrent network (1)