Gender Recognition Using a Gaze-Guided Self-Attention Mechanism Robust Against Background Bias in Training Samples

نویسندگان

چکیده

We propose an attention mechanism in deep learning networks for gender recognition using the gaze distribution of human observers when they judge people pedestrian images. Prevalent mechanisms spatially compute correlation among values all cells input feature map to calculate weights. If a large bias background images (e.g., test samples and training containing different backgrounds) is present, weights learned prevalent are affected by bias, which turn reduces accuracy recognition. To avoid this problem, we incorporate called gaze-guided self-attention (GSA) that inspired visual attention. Our method assigns suitable each observers. In particular, GSA yields promising results even with bias. The experiments on publicly available datasets confirm our GSA, distribution, more accurate than currently attention-based methods case between samples.

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

using game theory techniques in self-organizing maps training

شبکه خود سازمانده پرکاربردترین شبکه عصبی برای انجام خوشه بندی و کوانتیزه نمودن برداری است. از زمان معرفی این شبکه تاکنون، از این روش در مسائل مختلف در حوزه های گوناگون استفاده و توسعه ها و بهبودهای متعددی برای آن ارائه شده است. شبکه خودسازمانده از تعدادی سلول برای تخمین تابع توزیع الگوهای ورودی در فضای چندبعدی استفاده می کند. احتمال وجود سلول مرده مشکلی اساسی در الگوریتم شبکه خودسازمانده به حسا...

Voice-based Age and Gender Recognition using Training Generative Sparse Model

Abstract: Gender recognition and age detection are important problems in telephone speech processing to investigate the identity of an individual using voice characteristics. In this paper a new gender and age recognition system is introduced based on generative incoherent models learned using sparse non-negative matrix factorization and atom correction post-processing method. Similar to genera...

متن کامل

Understanding gender bias in face recognition: effects of divided attention at encoding.

Prior research has demonstrated a female own-gender bias in face recognition, with females better at recognizing female faces than male faces. We explored the basis for this effect by examining the effect of divided attention during encoding on females' and males' recognition of female and male faces. For female participants, divided attention impaired recognition performance for female faces t...

متن کامل

Attention Modifies Gender Differences in Face Recognition*

Gender differences favoring women have been found in face recognition, and in addition to this, it has been shown that women remember more female than male faces. This own-gender effect may be a result of women directing more attention towards female faces, resulting in a better memory. The aim of this study was to assess the role of attention for gender differences in face recognition and wome...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: IEICE Transactions on Information and Systems

سال: 2022

ISSN: ['0916-8532', '1745-1361']

DOI: https://doi.org/10.1587/transinf.2021edp7117