Publication Projects Experience Contact

Jiya Sinha

Incoming PhD student @ Perception and Intelligence Lab, IIT Kanpur

Previously, I completed my undergraduate studies in Data Science and Engineering at IISER Bhopal, where I worked under the guidance of Dr. Akshay Agarwal. My research interests lie in computer vision and audio processing, focusing on developing reliable, generalizable AI systems that perform well in real-world environments. I am looking forward to pursuing PhD @ IIT Kanpur, starting July 2026.


Download CV

Last updated: July 2026

Education

BS in Data Science and Engineering

2022 – 2026

IISER Bhopal

CGPA: 8.78/10

Research Areas
  • Multimodal Deep Learning (Vision/Audio/Language)
  • Robust AI
  • Applied Large Language Models
Selected Publications

Does Reality Matter? A Reality-Enhanced Audio SceneFake Dataset

Jiya Sinha, Aarthi S, A. Agarwal

IEEE International Joint Conference on Biometrics (IJCB 2026)

To be published

Gesture Recognition for Emergencies: Dataset and Cross-Condition Analysis

Jiya Sinha, P. Bhattacharya, A. Agarwal

IEEE International Conference on Automatic Face and Gesture Recognition (FG 2025)

Read Publication
Selected Projects

Does Reality Matter? A Reality-Aware Audio SceneFake Dataset

Bachelor's Thesis

Built a 15.42-hour reality-aware SceneFake audio dataset across 9 acoustic scenes, recorded under natural conditions with diverse speakers and environments. Demonstrates that existing deepfake detection models fail to generalize to realistic conditions, highlighting the need for real-world data.

View Thesis

Graph-Based Makeup Transfer using GCNs

Structured facial graph-based makeup transfer using GCNs, evaluated through a user study with perceptual scoring.

View Project

Face Recognition using classical CV techniques

Built a face recognition pipeline using the Yale Face Database with hand-crafted features and classical ML models.

View Project
Experience

Research Intern

Trustworthy BiometraVision Lab, IISER Bhopal

Sep 2024 – Jan 2025

  • Built a 1,000+ video gesture recognition dataset across diverse environments and devices.
  • Benchmarked models including ViViT, VGG-LSTM, and MobileNetV2-LSTM.
  • Co-authored a paper accepted at IEEE FG 2025.

Research Intern

Advanced Signal and Image Processing Lab (ASIP), IISER Bhopal

Mar 2025 – Aug 2025

  • Designed deep learning architectures for NIR-to-RGB image colorization.
  • Explored GAN-based generator variations and perceptual loss formulations.
  • Optimized for perceptual quality.