Search

Your search keyword '"Jia, Zhijie"' showing total 321 results
Start Over Author "Jia, Zhijie"
321 results on '"Jia, Zhijie"'

2. ViT-Calibrator: Decision Stream Calibration for Vision Transformer

Author

Chen, Lin, Jia, Zhijie, Qiu, Tian, Cheng, Lechao, Lei, Jie, Feng, Zunlei, and Song, Mingli

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

A surge of interest has emerged in utilizing Transformers in diverse vision tasks owing to its formidable performance. However, existing approaches primarily focus on optimizing internal model architecture designs that often entail significant trial and error with high burdens. In this work, we propose a new paradigm dubbed Decision Stream Calibration that boosts the performance of general Vision Transformers. To achieve this, we shed light on the information propagation mechanism in the learning procedure by exploring the correlation between different tokens and the relevance coefficient of multiple dimensions. Upon further analysis, it was discovered that 1) the final decision is associated with tokens of foreground targets, while token features of foreground target will be transmitted into the next layer as much as possible, and the useless token features of background area will be eliminated gradually in the forward propagation. 2) Each category is solely associated with specific sparse dimensions in the tokens. Based on the discoveries mentioned above, we designed a two-stage calibration scheme, namely ViT-Calibrator, including token propagation calibration stage and dimension propagation calibration stage. Extensive experiments on commonly used datasets show that the proposed approach can achieve promising results. The source codes are given in the supplements., Comment: At present, the paper involves internal projects of the company, and it is not convenient to publish it temporarily, so the article needs to be withdrawn temporarily

Published
2023

4. Model Doctor for Diagnosing and Treating Segmentation Error

Author

Jia, Zhijie, Chen, Lin, Hu, Kaiwen, Cheng, Lechao, Feng, Zunlei, and Song, Mingli

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Despite the remarkable progress in semantic segmentation tasks with the advancement of deep neural networks, existing U-shaped hierarchical typical segmentation networks still suffer from local misclassification of categories and inaccurate target boundaries. In an effort to alleviate this issue, we propose a Model Doctor for semantic segmentation problems. The Model Doctor is designed to diagnose the aforementioned problems in existing pre-trained models and treat them without introducing additional data, with the goal of refining the parameters to achieve better performance. Extensive experiments on several benchmark datasets demonstrate the effectiveness of our method. Code is available at \url{https://github.com/zhijiejia/SegDoctor}.

Published
2023

45. Synthesis and characterization of a high-strength alumina ceramic reinforced by AlN-Al2O3 coating.

Author

Fu, Shuai, Jia, Zhijie, Ding, Wei, Bao, Yiwang, and Wan, Detian

Subjects
*ALUMINUM nitride, *CERAMIC coating, *SUBSTRATES (Materials science), *THERMAL expansion, *CRACK propagation (Fracture mechanics)
Abstract

In this study, a high-strength alumina ceramic reinforced by a coating composed of 25wt% AlN and 75 wt% Al2O3 (labeled as ACRA-CoA) was fabricated using the spark plasma sintering (SPS) method. The thermal expansion coefficient difference between the coating and the substrate generates significant residual compressive stress in the coating during sintering cooling process, which acts protectively on the ceramic substrate, hindering crack propagation and enhancing mechanical strength. As a result, ACRA-CoA demonstrates an impressive flexural strength of 586.4 ± 70.5 MPa, marking a substantial 65.7% increase compared to conventional alumina's strength of 353.9 ± 20.5 MPa. Additionally, the doped aluminium nitride in the coating significantly improves its thermal properties, for example, at 25 °C, the thermal conductivity of ACRA-CoA is 28.93 W∙m − 1 ∙K − 1, which is 46.2% higher than that of conventional alumina, and the thermal diffusion coefficient increased by 43.5% to 0.089 mm − 2 ∙S − 1, while the coefficient of thermal expansion is 8.5 × 10 − 6 K − 1 [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results