Published August 18, 2026 | Version v2

FER2025: Explainable Binary Gender-Aware Facial Emotion Detection via Hybrid Deep Learning and Customized Visual Dataset

  • 1. Student
  • 2. Bsc. in Dept. of CSE, United International University (UIU)

Description

This dataset supports facial emotion recognition projects and includes gender classification features. The metadata maps 7,386 images containing 11,253 facial annotations sourced from Unsplash and commercial platforms. Due to strict commercial copyright terms, raw image files are withheld, and we publicly share the comprehensive metadata schema instead. The dataset features 12 distinct classes covering six primary emotions across male and female subjects, split into training (70%), validation (20%), and testing (10%) sets. The file metadata_FER2025.csv structures this data by detailing bounding boxes, class designations, and subset splits to enable efficient object detection training.

Methods

To reproduce the dataset preparation and model development process, researchers must first independently collect facial images from appropriate third-party sources in accordance with the applicable licensing terms. The collected images should then be uploaded to Roboflow, where facial expressions are annotated using bounding boxes and assigned to the corresponding 12 class labels representing six emotion categories across two gender groups. After annotation, the dataset can be organized into training, validation, and test subsets using the designated 70/20/10 split. Standard preprocessing procedures, including image resizing, normalization, and optional data augmentation, may then be applied. Finally, the annotated dataset can be exported in a suitable format, imported into the selected deep learning framework, and used to train and evaluate the proposed or other deep learning models on the designated validation and test sets.

Files

metadata_FER2025.csv

Files (1.4 MB)

Name Size
md5:f399276de746a1ac2285785e5a67828c
1.4 MB Preview Download

Additional details

Dates

Available
2026-04-21