553 citations · 3.2k across the 83 of their papers we have counts for
6 papers · 1 filter
Efficient and Invariant Convolutional Neural Networks for Dense Prediction
Hongyang Gao, Shuiwang Ji
Convolutional neural networks have shown great success on feature extraction from raw input data such as images. Although convolutional neural networks are invariant to translation…
Dense Transformer Networks
Jun Li, Yongjun Chen, Lei Cai +2
The key idea of current deep learning methods for dense prediction is to apply a model on a regular patch centered on each pixel to make pixel-wise predictions. These methods are l…
Multi-Stage Variational Auto-Encoders for Coarse-to-Fine Image Generation
Lei Cai, Hongyang Gao, Shuiwang Ji
Variational auto-encoder (VAE) is a powerful unsupervised learning framework for image generation. One drawback of VAE is that it generates blurry images due to its Gaussianity ass…
Learning Convolutional Text Representations for Visual Question Answering
Zhengyang Wang, Shuiwang Ji
Visual question answering is a recently proposed artificial intelligence task that requires a deep understanding of both images and texts. In deep learning, images are typically mo…
Spatial Variational Auto-Encoding via Matrix-Variate Normal Distributions
Zhengyang Wang, Hao Yuan, Shuiwang Ji
The key idea of variational auto-encoders (VAEs) resembles that of traditional auto-encoder models in which spatial information is supposed to be explicitly encoded in the latent s…
Pixel Deconvolutional Networks
Hongyang Gao, Hao Yuan, Zhengyang Wang +1
Deconvolutional layers have been widely used in a variety of deep models for up-sampling, including encoder-decoder networks for semantic segmentation and deep generative models fo…