近期智能工程系邀请了美国约翰霍普金斯大学的Trac D Tran教授开展系列学术讲座“稀疏性、神经网络和成像应用的最新进展系列报告”，通过腾讯直播的方式进行。

 

Title: 

Sparsity, Neural Networks and Imaging Applications: Recent Advances. 

 

Abstract:

This lecture series presents several recent advances in sparse coding, 

dictionary learning, their connections to deep neural networks, and their

applications in image/video processing. The series covers five related topics:

1. Compressive hardware systems for high-dimensional data acquisition

2. Entropy minimization for sparse recovery, low-rank approximation and representative selection

3. Deep sparse coding networks versus CNN 

4. Neural networks for beamforming in ultrasound and radar applications 

5. EffiScene: efficient rigidity inference for unsupervised joint learning of optical

flow, depth, camera pose and motion segmentation.

These works are in collaboration with Prof. Guangming Shi, Prof. Mark Foster, Prof. Muyinatu A. Lediju Bell, Prof. Sang Peter Chin, Jaewook Shin, Dung N. Tran, Shuai Huang, Arun Nair, Minh Bui, and Yang Jiao.