Current research in image processing involves the investigation of the use of sparse representations in image coding, denoising and compressive sensing. In addition, algorithms for 2D to 3D image transformations are also developed. Research in video processing and analysis includes the design of strategies for energy-efficient transmission, distributed coding and compressive acquisition. Furthermore, implementation aspects using multi-core DSP platforms are also considered.

• Multilevel Dictionary Learning for Sparse Representation of Images
• Compressive Acquisition of Dynamic Scenes
• Transform-domain distributed video coding with rate–distortion-based adaptive quantization
• Objective Video Quality Assessment Methods: A Classification, Review, and Performance Comparison
• Trends in Multi-Core DSP Platforms
• Increased depth perception with sharpness enhancement for stereo video

Current research in machine learning/computer vision at the SenSIP Center includes developing sparse representation based frameworks for object recognition and image retrieval, and investigating theoretical characteristics/performance of unsupervised clustering algorithms. In addition, efficient implementation of image retrieval systems using Graphical Processing Units and smartphones are performed.

• Learning Dictionaries for Local Sparse Coding in Image Classification
• Implementation of Fast Image Coding and Retrieval using GPU
• Optimality and Stability of K-hyperline Clustering