Bhavan Jasani

Applied Scientist · Computer Vision & Multimodal AI

I build practical machine learning systems at the intersection of vision, language, and reasoning. My work spans document intelligence, visual grounding, and chart VQA—shipping research to production.

Portrait of Bhavan Jasani

About

I'm an Applied Scientist II focused on computer vision and multimodal learning. My work bridges research and product—turning ideas into shipped features like document VQA and multimodal assistants. Increasingly, I'm motivated to apply AI to healthcare‑impactful problems that improve patient outcomes. Areas: document intelligence (layout‑aware transformers), visual grounding, chart reasoning/VQA, synthetic data generation, and efficient training/inference at scale.

Experience

  1. Amazon AWS AI Labs

    Amazon AWS AI Labs

    Applied Scientist II

    Sep 2019 – Present

  2. Carnegie Mellon University

    Carnegie Mellon University, Robotics Institute

    Research Assistant

    Oct 2017 – Aug 2019

  3. Nanyang Technological University

    Nanyang Technological University, Singapore

    Research Assistant

    May 2016 – Aug 2016

Selected Publications

  1. Chart VQA

    Synthesize Step-by-Step: Tools, Templates and LLMs as Data Generators for Reasoning-Based Chart VQA

    CVPR 2024. Li, Jasani, Tang, Ghadar.

    PDF
  2. YORO

    YORO: Lightweight End-to-End Visual Grounding

    ECCV 2022 Workshops. Ho, Appalaraju, Jasani, Manmatha, Vasconcelos.

    PDF
  3. DocFormer

    DocFormer: End-to-End Transformer for Document Understanding

    ICCV 2021. Appalaraju, Jasani, Kota, Xie, Manmatha.

    PDF
  4. MovieQA

    Are We Asking the Right Questions in MovieQA?

    ICCV 2019 Workshop. Bhavan Jasani et al.

    PDF
  5. CMU Thesis

    Automatic detection of human affective behavior in dyadic conversations

    CMU RI Technical Report (Master's Thesis), 2019. Bhavan Jasani.

    PDF
  6. Pose Action Recognition

    Skeleton-based Zero Shot Action Recognition in Joint Pose-Language Semantic Space

    Workshop 2019. Bhavan Jasani et al.

    PDF
  7. T-CSVT Harris Corner

    Threshold-Guided Design and Optimization for Harris Corner Detector Architecture

    IEEE T-CSVT, 2017. Jasani, Lam, Meher, Wu.

    PDF

Curriculum Vitae

Download my latest CV (updated 2025):

Contact

Email is best. For collaborations, please include a brief summary and relevant links.