Dataset Containing Frame Images of Indian Candidates during Interview for Confidence Level Estimation

Published: 14 March 2023| Version 2 | DOI: 10.17632/dk2h3vhnnx.2
, Nishigandha Satav, Abhishek Kakade, Rutuja Shastri


The dataset contains a set of 4754 images of Indian candidates during interviews, which can be used for various machine learning and computer vision applications, such as face recognition, emotion detection, eye moment, face direction, and gender recognition. The dataset contains more than 170 images of each Indian candidate during interviews, captured in various settings, including indoor and outdoor locations, with different lighting conditions and backgrounds. Each image is in JPEG format. Furthermore, each candidate is denoted by his/her id number, in total there are 21 candidates with their labeled confidence (1)/ unconfidence(0). The dataset is labeled considering various features of the candidates, such as face direction, facial expression, and eye moment. These labels can be used for supervised learning tasks, such as classification or regression. The dataset is diverse in terms of the gender and age of the candidates, making it suitable for training machine learning models.



Savitribai Phule Pune University


Artificial Intelligence, Computer Vision, Facial Recognition, Eyes Recognition, Frame, Interviewing