135-class Emotional Facial Expression Dataset

The ability to perceive human facial emotions is an essential feature of various multi-modal applications, especially in the intelligent human-computer interaction (HCI) area. In recent decades, considerable efforts have been put into researching automatic facial emotion recognition (FER). However, most of the existing FER methods only focus on either basic emotions such as the seven/eight categories (e.g., happiness, anger and surprise) or abstract dimensions (valence, arousal, etc.), while neglecting the fruitful nature of emotion statements. In real-world scenarios, there is definitely a larger vocabulary for describing human inner feelings as well as their reflection on facial expressions. This dataset addresses the semantic richness issue in the FER problem, with an emphasis on the granularity of the emotion concepts. Particularly, we take inspiration from former psycho-linguistic research, which conducted a prototypicality rating study and chose 135 emotion names from hundreds of English emotion terms.Based on the 135 emotion categories, the dataset collects a large-scale 135-class FER image dataset. The paper [1] demonstrates the accessibility of prompting FER research to a fine-grained level by conducting extensive evaluations on the dataset credibility and the accompanying baseline classification model. To the best of our knowledge, this is the first dataset aimed at exploiting such a large semantic space for emotion representation in the FER problem.[1] K. Chen, X. Yang, C. Fan, W. Zhang and Y. Ding, "Semantic-Rich Facial Emotional Expression Recognition," in IEEE Transactions on Affective Computing, vol. 13, no. 4, pp. 1906-1916, 1 Oct.-Dec. 2022, doi: 10.1109/TAFFC.2022.3201290.

感知人类面部情绪的能力是诸多多模态应用的核心特征，尤其在智能人机交互（Human-Computer Interaction, HCI）领域。近数十年来，学界已在自动面部表情识别（Facial Emotion Recognition, FER）方向投入了大量研究精力。然而，现有多数FER方法仅聚焦于七/八大类基础情绪（如快乐、愤怒、惊讶等）或抽象情绪维度（效价、唤醒度等），却忽略了情绪表述所承载的丰富语义内涵。在真实应用场景中，用于描述人类内心感受及其面部表情映射的词汇量显然更为庞大。本数据集旨在解决FER任务中语义丰富度不足的问题，着重关注情绪概念的细粒度划分。具体而言，我们借鉴了此前的心理语言学研究——该研究开展了典型性评级实验，从数百个英语情绪术语中筛选出135个情绪名称。基于这135个情绪类别，本数据集构建了一个大规模的135类FER图像数据集。文献[1]通过全面评估数据集可信度并搭建配套基线分类模型，验证了将FER研究推进至细粒度层面的可行性。据我们所知，本数据集是首个针对FER任务中此类大规模情绪表征语义空间的专用数据集。
[1] K. Chen、X. Yang、C. Fan、W. Zhang 与 Y. Ding，《语义丰富型面部情绪表达识别》，载于《IEEE情感计算汇刊》，第13卷第4期，第1906-1916页，2022年10月-12月，DOI: 10.1109/TAFFC.2022.3201290。

搜集汇总

背景与挑战

背景概述

135-class Emotional Facial Expression Dataset是一个大规模面部表情识别数据集，包含135种情感类别和696,168张面部图像，旨在推动细粒度的情感识别研究。数据集分为训练、验证和测试三部分，分别包含556,803、69,560和69,805张图像，每张图像都标注了情感类别和详细描述。