Download full-text PDF |
Source |
---|---|
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2696863 | PMC |
http://dx.doi.org/10.3238/arztebl.2008.0293a | DOI Listing |
Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit
June 2024
Fine-tuning pre-trained vision-language models, like CLIP, has yielded success on diverse downstream tasks. However, several pain points persist for this paradigm: (i) directly tuning entire pre-trained models becomes both time-intensive and computationally costly. Additionally, these tuned models tend to become highly specialized, limiting their practicality for real-world deployment; (ii) recent studies indicate that pre-trained vision-language classifiers may overly depend on spurious features - patterns that correlate with the target in training data, but are not related to the true labeling function; and (iii) existing studies on mitigating the reliance on spurious features, largely based on the assumption that we can identify such features, does not provide definitive assurance for real-world applications.
View Article and Find Full Text PDFExp Mol Med
November 2024
Division of Thyroid Surgery, Department of General Surgery, West China Hospital, Sichuan University, Chengdu, China.
PLoS One
November 2024
Faculty of Information Engineering, Xinjiang Institute of Technology, Aksu, China.
The objective of image fusion is to integrate complementary features from source images to better cater to the needs of human and machine vision. However, existing image fusion algorithms predominantly focus on enhancing the visual appeal of the fused image for human perception, often neglecting their impact on subsequent high-level visual tasks, particularly the processing of semantic information. Moreover, these fusion methods that incorporate downstream tasks tend to be overly complex and computationally intensive, which is not conducive to practical applications.
View Article and Find Full Text PDFAnn Appl Stat
December 2023
Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison.
Individualized treatment rules (ITRs) for treatment recommendation is an important topic for precision medicine as not all beneficial treatments work well for all individuals. Interpretability is a desirable property of ITRs, as it helps practitioners make sense of treatment decisions, yet there is a need for ITRs to be flexible to effectively model complex biomedical data for treatment decision making. Many ITR approaches either focus on linear ITRs, which may perform poorly when true optimal ITRs are nonlinear, or black-box nonlinear ITRs, which may be hard to interpret and can be overly complex.
View Article and Find Full Text PDFInt J Mol Sci
August 2024
Institute of Germplasm Resources and Biotechnology, Tianjin Academy of Agricultural Sciences, Tianjin 300192, China.
Soil salinization severely limits the quality and productivity of economic crops, threatening global food security. Recent advancements have improved our understanding of how plants perceive, signal, and respond to salt stress. The discovery of the Salt Overly Sensitive (SOS) pathway has been crucial in revealing the molecular mechanisms behind plant salinity tolerance.
View Article and Find Full Text PDFEnter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!