Facing the direct impact of video quality on viewing experience, especially regarding the presentation of facial details, current methods such as simply applying super-resolution networks to face datasets or independently processing each frame often struggle to balance the detail of facial reconstruction with temporal consistency. To address this challenge, the research team at Nanyang Technological University has launched the innovative framework KEEP, which utilizes the principles of Kalman filtering to achieve a 'memory' capability in facial restoration, significantly enhancing the consistency and continuity of facial details during the restoration process. The KEEP framework consists of four key modules: Encoder, Decoder, Kalman.

The research team led by Yuan Li at Peking University, in collaboration with researchers from Nanyang Technological University and Tsinghua University, has developed an innovative image retrieval technology that aims to precisely match user needs through sketches, artworks, and even low-resolution images. The core of this technology is the 'Universal Style Retrieval' method, which can handle multiple query styles including combinations of text, sketches, and artworks, significantly enhancing retrieval flexibility and accuracy. The research team constructed the DSR and ImageNet-X datasets to provide training and testing resources for the new method.

The team from Nanyang Technological University has launched the InsActor framework, which utilizes a diffusion-based human motion model to generate realistic physical animations driven by instructions. InsActor achieves conditional motion planning through a diffusion strategy, capturing the complex relationship between high-level human instructions and character actions. Experiments demonstrate that InsActor achieves advanced results in instruction-driven motion generation and waypoint-oriented tasks. The framework's flexibility allows for customizable animations, showcasing wide applicability while maintaining visual appeal and physical feasibility.