Dataset of Indian face images with various expressions

Fast and precise human emotion classification and detection is a quality parameter in various industries and research areas. This is a continuously evolving research area that consistently garners attention from researchers. Machine Learning (ML), deep learning, and Artificial Intelligence (AI) methods have demonstrated considerable potential in solving classification and object detection challenges. A pristine and well-curated dataset stands as the fundamental prerequisite for constructing precise and resilient ML and AI models suitable for real-time applications. With this objective in mind, compiled an image dataset consisting of Indian facial expressions with Happiness, Thoughtfulness, Anger, Surprise, Sadness, Fear, and Excited and these facial expressions are captured in three directions front, left side, and right side. The dataset has been segregated into six folders: 1) Happiness 2) Thoughtfulness 3) Anger 4) Surprise 5) Sadness 6) Fear 7) Excited. The dataset comprises over 6,000+ images in raw format. We believe that this raw dataset will significantly aid in training, testing, and validating machine learning models for facial expression classification or recognition.

快速且精准的人类情感分类与检测，是诸多行业与研究领域中的关键质量参数。该领域作为持续演进的研究方向，始终受到研究者们的广泛关注。机器学习（Machine Learning，ML）、深度学习以及人工智能（Artificial Intelligence，AI）相关方法在解决分类与目标检测难题方面展现出了可观的应用潜力。一套原始且经过精心整理的数据集，是构建适用于实时应用场景的精准、鲁棒的机器学习与人工智能模型的核心先决条件。为此，我们构建了一套涵盖印度人七种面部表情的图像数据集，具体包括快乐、沉思、愤怒、惊讶、悲伤、恐惧与兴奋，且所有表情样本均从正面、左侧面、右侧面三个拍摄角度采集得到。本数据集共分为七个文件夹，依次为：1）快乐 2）沉思 3）愤怒 4）惊讶 5）悲伤 6）恐惧 7）兴奋。该数据集包含超过6000张原始格式图像。我们认为，这套原始数据集将能够为面向面部表情分类或识别任务的机器学习模型的训练、测试与验证提供显著助力。