纠正视觉语言模型中的表示偏差,以实现长尾识别.
Bo Li1, Yongqiang Yao2, Jingru Tan3
1Tongji University, No. 4800 Caoan Road, Shanghai, 201804, China.
概括
这项研究通过考虑类相关性来解决长尾动物视觉识别表现不佳的问题. 一种新的方法,纠正对比术语 (ReCT),减少了对象学习中的偏差,以获得更好的准确性.
科学领域:
- 计算机科学 计算机科学
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 自然数据通常具有长尾分布,导致罕见类的识别性能差.
- 现有的方法主要集中在类频率上,忽视了类相关性的关键因素.
研究的目的:
- 在视觉语言 (VL) 框架内调查长尾视觉识别的性能瓶.
- 提出一种新的方法,将类相关性纳入其中,以解决头和尾类之间的识别混.
主要方法:
- 将表示学习建模成特殊和共同的部分,以捕捉独特和共享的类特征.
- 引入纠正对比术语 (ReCT) 以减轻共同表示学习中的偏见.
- 利用语义提示和培训状态来指导纠正过程.
主要成果:
- 共同的代表性学习偏向于头类,导致网络优先考虑共享的特征而不是独特的特征.
- ReCT有效地纠正了表示偏差,提高了尾部类的识别精度.
- 三个长尾数据集的实验表明,ReCT提高了性能,在iNaturalist2018上达到75.4%的准确性,使用ResNet-50骨干.
结论:
- 类相关性是长尾动物视觉识别的关键因素,与类频率一起.
- 拟议的ReCT方法为VL框架中的代表性偏见提供了有效的解决方案.
- 这项工作通过提高准确性和减少相关类之间的混来推进长尾识别的最新技术.
相关概念视频
Improving Translational Accuracy
10.5K
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
10.5K
Vision
53.4K
Vision is the result of light being detected and transduced into neural signals by the retina of the eye. This information is then further analyzed and interpreted by the brain. First, light enters the front of the eye and is focused by the cornea and lens onto the retina—a thin sheet of neural tissue lining the back of the eye. Because of refraction through the convex lens of the eye, images are projected onto the retina upside-down and reversed.
53.4K
Residuals and Least-Squares Property
7.4K
The vertical distance between the actual value of y and the estimated value of y. In other words, it measures the vertical distance between the actual data point and the predicted point on the line
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
7.4K
Bias
4.2K
Bias refers to any tendency that prevents a question from being considered unprejudiced. In research, bias occurs when one outcome or answer is selected or encouraged over others in sampling or testing. Bias can occur during any research phase, including study design, data collection, analysis, and publication.
In statistics, a sampling bias is created when a sample is collected from a population, and some members of the population are not as likely to be chosen as others (remember, each member...
In statistics, a sampling bias is created when a sample is collected from a population, and some members of the population are not as likely to be chosen as others (remember, each member...
4.2K
Difference from Background: Limit of Detection
6.4K
The limit of detection (LOD) is the smallest amount of analyte that can be distinguished from the background noise. The LOD value corresponds to the concentration at which the analyte signal is three times larger than the standard deviation of the blank signal. Below this value, the analyte signal cannot be differentiated from the background noise. It is calculated by dividing the calibration slope by 3 times the standard deviation of the blank signals.
The LOD indicates the presence or absence...
The LOD indicates the presence or absence...
6.4K
Lateralization
336
Brain lateralization refers to the division of mental processes and functions between the two hemispheres of the brain, a phenomenon that optimizes neural efficiency and underpins complex abilities in humans. This specialization allows each hemisphere to perform tasks where it has a comparative advantage, facilitating more refined cognitive capabilities across different domains.
336


