Skip to content
View xiaofengShi's full-sized avatar
🎯
Focusing
🎯
Focusing
376 followers · 11 following

Achievements

Achievement: Pull SharkAchievement: Arctic Code Vault ContributorAchievement: Starstruckx3

Achievements

Achievement: Pull SharkAchievement: Arctic Code Vault ContributorAchievement: Starstruckx3

Block or report xiaofengShi

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
xiaofengShi/README.md

AI Researcher & Engineer @ BAAI

I am an AI researcher and engineer interested in making foundation models useful in specialized domains. My work spans open datasets, domain adaptation, post-training and reinforcement learning, agentic research systems, and multimodal reasoning.

My path has taken me from ByteDance and Meituan to BAAI, where I build end-to-end systems from data and evaluation to models and applications. I care about practical, reproducible research, and I try to release the code, datasets, and tools behind my work whenever possible.

Blog | Google Scholar | Hugging Face | Patents | Email

Selected Research

My full paper list is available on my research page. I have published papers at ICML, AAAI, and COLING.

Earlier Open Source

  • CHINESE-OCR Stars: end-to-end natural-scene Chinese text detection and recognition with CTPN, CRNN, and CTC
  • Image2Katex Stars: image-to-LaTeX recognition for printed and handwritten mathematical formulas
  • DKT-TensorFlow Stars: Deep Knowledge Tracing for adaptive learning

Pinned Loading

  1. CHINESE-OCR CHINESE-OCR Public

    End-to-end Chinese scene-text detection and recognition with CTPN, CRNN, and CTC (legacy project).

    Python 3k 941

  2. SPAR SPAR Public

    SPAR: multi-agent scholarly retrieval with query decomposition, query evolution, and citation-aware exploration.

    Python 28 8

  3. MechVQA MechVQA Public

    ICML 2026 benchmark and MechVL models for multimodal understanding of mechanical engineering drawings.

    Python 24 2

  4. Image2Katex Image2Katex Public

    Image-to-LaTeX recognition for printed and handwritten mathematical formulas.

    HTML 293 73