I’m Abhinav Srivastava, B.Tech CSE (AI specialization) student at Bennett University. My journey spans traditional ML and computer vision to deep learning, transformers, and multimodal AI systems. Currently exploring Generative AI and efficient vision transformers to build accurate, explainable, and real-world solutions.
Machine Learning & Deep Learning Intern – AIQUANTUM Smart Solutions Pvt. Ltd. (May 2024 – July 2024)
-> Designed and optimized deep learning models for medical imaging and satellite vision tasks.
-> Achieved 97.5% accuracy in liver cancer histopathology classification using a custom ResNet-50, setting a dataset benchmark.
-> Worked with TensorFlow, OpenCV, Python on end-to-end AI pipelines.
1. DisasterFormer: An Enhanced Transformer with Edge-Guided Attention for Disaster Image Recognition - Published at
IATMSI-2025 (Conference Paper) | DOI: https://doi.org/10.1109/IATMSI64286.2025.10985468
2. TriVerseNet: Multi-Branch ResNet Framework for Enhanced Endoscopic Lesion Classification via Edge and Texture
Integration - Published at NETCRYPT-25 (Conference Paper) | DOI: https://doi.org/10.1109/NETCRYPT65877.2025.11102233
3. Small-Scale ConvNeXt Variant Optimized for Pneumonia Classification on Chest X-Ray Images – Published at
IACIS-2025 (Conference Paper) | DOI: https://doi.org/10.1109/IACIS65746.2025.11211139
4. EffiCoordNet: Towards Lightweight and Accurate Violence Detection from Single Video Frames Using Coordinate-Aware Networks – Published at CISES-2025 (Conference Paper) | DOI: https://doi.org/10.1109/CISES66934.2025.11265571
-> Delivered 97.5% accuracy benchmark in liver histopathology classification during internship.
-> Reached Top 50 in HackWars Hackathon (out of 1500+ teams).
-> Project recognized for innovation in HackCBS 7.0 (Vultr Track).
-> In 2025, I submitted three peer-reviewed conference papers and am currently engaged in several additional research projects, with a greater emphasis on journal publication.
-> Solved 100+ DSA/CP problems across LeetCode, GFG, HackerRank.