Speaker: Professor Wu Jinjian

Date: November 10, 2020

Time: 9:00

Venue: Meeting Room B318, Building 5, South Campus

Intro to the lecture:

High-quality image data is an important means for us to understand the world. Facing the challenge of long-distance, high-speed movement, and obscured target detection, how to be able to see clearly, grasp, and distinguish clearly is a huge challenge. In response to this problem, the research team has carried out research in three directions: simulative imaging, intelligent evaluation, and brain-like recognition around the three stages of perception, evaluation, and knowledge. Sense: Subverting the traditional imaging mode based on energy integration, designing a bionic photosensitive chip based on energy difference, breaking the bottleneck of imaging dynamic range and imaging rate. Evaluation: Revealing the internal causes of quality attenuation, developing a perceptual purpose-oriented objective quality evaluation method, and solving the problem in measuring attenuation strength. Knowing: Learning from the cognitive mechanism of the human brain, realizing small target recognition based on knowledge embedding, attention guidance, and scene constraints, which has been used in satellite remote sensing and has effectively improved its space reconnaissance capabilities.