National Innovation Project

Guided by Professor Zuo Wangmeng.

The project primarily focuses on creatively modifying artworks. It leverages the latent space of images for editing, enabling the transformation of static images into continuous motion videos.

Masked Unmasked Face Recognition

NUS Summer Workshop

The project employs techniques of random occlusion to enhance the model's ability to recognize and detect faces under a range of conditions where the face may be partially obscured.

Battery Defect Detection

Sichuan Changhong Electric Co.,Ltd. AI laboratory

Various data augmentation methods are used to overcome the scarcity of industrial defect data, enhancing the performance of segmentation and detection.

MaeFuse

Guided by Professor Jiang Junjun, completed as the first author.

Our model utilizes a pretrained encoder from Masked Autoencoders (MAE), which facilities the omni features extraction for low-level reconstruction and high-level vision tasks, to obtain perception friendly features with a low cost. In order to eliminate the domain gap of different modal features and the block effect caused by the MAE encoder, we further develop a guided training strategy. This strategy is meticulously crafted to ensure that the fusion layer seamlessly adjusts to the feature space of the encoder, gradually enhancing the fusion effect. It facilitates the comprehensive integration of feature vectors from both infrared and visible modalities, preserving the rich details inherent in each. MaeFuse not only introduces a novel perspective in the realm of fusion techniques but also stands out with impressive performance across various public datasets.

[code] [pdf]