• Meta Learnning

Goal of meta-learning is to learn a learning algorithm that is flexible and can solve tasks given at test-time. It is an emerging and fast-developing machine learning research topic, with various computer vision and machine learning applications.

• Semi-supervised Learning

Semi-supervised learning aims to learn from both labeled and unlabeled data. In addition to supervised learning with labeled data, semi-supervised learning utilizes the underlying structure of the data distribution.

Video understanding is a research field with practical computer vision tasks, including human action recognition, video object detection and segmentation, video summarization, etc. With a large volume of video data available online, it is a prospective research field with a wide range of applications.

• Visual Tracking

Visual tracking aims to track and locate the target object inside consecutive video frames, given the initial target state. Its applications include autonomous driving, automated surveillance, and robotics.

Goal of image restoration is to restore or enhance the corrupted image, and it includes sub-problems of super-resolution, deblurring, denoising, etc. It is a challenging problem with various camera-related applications.

• Super-Resolution

Super-resolution aims to enhance the resolution of an image beyond its original resolution. Given low-resolution image, the goal is to generate a high-resolution version of an image using additional information and prior knowledge. 

• Video Frame Interpolation

Video frame interpolation algorithms can generate new frames between existing frames of a video sequence, where the are commonly used to increase the frame rate of a video for smoother motion and improved visual quality.