Scholarlink

In the field of computer vision, automatically analyzing these unsaid feelings has proven to be an intriguing and difficult problem. Its potential uses span a variety of domains, such as psychology, process automation, and product marketing. These wasn’t an easy undertaking, since people vary a great deal in how they convey their feelings using their facial expressions. Computer vision is one of the many domains of research that has seen significant development thanks in large part to machine learning, and more specifically deep learning. We address the issue of facial sentiment analysis in this research study by introducing an architecture implemented with convolutional neural networks (CNNs). FER-2013 public dataset has been used by us for training and testing. This work has been done in stages: first, the data has been preprocessed; next, features have been extracted; and last, our trained model network has been used to classify the data. Our experiment’s results are quite promising and represent a breakthrough in the field of automated facial sentiment analysis.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An Implementation of Convolutional Neural Network-Based Architecture to Address Facial Sentiment Analysis

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • M. Ravindran,
  • G. Madhukar,
  • A. Srinivasula Reddy,
  • Mandapati Raja,
  • R. Venkateswara Reddy,
  • R. Suhasini

摘要

In the field of computer vision, automatically analyzing these unsaid feelings has proven to be an intriguing and difficult problem. Its potential uses span a variety of domains, such as psychology, process automation, and product marketing. These wasn’t an easy undertaking, since people vary a great deal in how they convey their feelings using their facial expressions. Computer vision is one of the many domains of research that has seen significant development thanks in large part to machine learning, and more specifically deep learning. We address the issue of facial sentiment analysis in this research study by introducing an architecture implemented with convolutional neural networks (CNNs). FER-2013 public dataset has been used by us for training and testing. This work has been done in stages: first, the data has been preprocessed; next, features have been extracted; and last, our trained model network has been used to classify the data. Our experiment’s results are quite promising and represent a breakthrough in the field of automated facial sentiment analysis.