Wuhan University · Software Engineering · 2027
Start from the problem itself
I am Bohan Su, an undergraduate in Software Engineering at Wuhan University. My current work spans visual representation under appearance variation, small-model structured scientific reasoning, and multi-agent systems for real tasks. I care about methods that become reproducible, analyzable, and deployable systems rather than isolated benchmark gains.
Research State
Three active research tracks
Appearance-variation vision, schema-first scientific reasoning, and multi-agent systems are moving in parallel as one coherent portfolio.
Current Output
Papers, systems, and competitions together
An ECCV 2026 submission, PEARL system progress toward AAAI, a TPAMI survey draft, MCM 2026 Finalist, and national-level systems work define the current output.
Collaboration Fit
Best for reproducible, deployable work
I contribute most on research prototypes and systems projects that need explicit evaluation logic, engineering closure, and clear writing.
Recent Signals
Recent Updates
Only actual research, systems, and competition progress is kept here.
TPAMI survey main body drafted
The main survey structure is stable; current effort is on tightening the introduction, wording, and the boundaries of the argument.
ECCV 2026 work moved into review
A submission on Universal ReID under appearance variation has entered review; the public layer now keeps the problem statement and technical shape without exposing the full package.
PEARL pipeline ready; targeting AAAI
The core implementation and verification-repair loop are in place; the next phase is main-experiment execution, attribution analysis, and tighter paper framing for AAAI submission.
MCM 2026 Finalist Award
SmartPhone Battery modeling project earned Finalist recognition with dual-branch ECM-EKF + PINN approach achieving 1.24% SOC RMSE and 8.7% TTE MAPE.
A2A multi-agent system delivered
Built a Holos + A2A workflow with architectural agents, task agents, protocol adaptation, streaming, fallback, and observability for end-to-end demos.
Current Focus
What matters most right now
- Translating research claims into evaluation protocols that others can actually audit.
- Building systems with diagnosis, repair, and observability rather than only a happy-path demo.
- Finding problem structures that transfer between vision tasks and reasoning systems.
Main Tracks
Research Directions and Papers
I am most interested in topics that connect method design, system behavior, and evaluation logic.
Direction 01
Visual representation under appearance variation
Studying how visual evidence breaks and reorganizes under occlusion, clothing change, cross-modality, and in-the-wild viewpoints in Universal ReID, including UniGeo's reliability-aware 3D geometric injection and a broader survey-level taxonomy.
Direction 02
Schema-first structured scientific reasoning with small models
Decoupling generation from verification with training-free curation and student SFT. PEARL positions as a complete framework for scientific reasoning graph extraction, targeting small models like Qwen3.5-4B with curriculum learning and verifiable supervision.
Direction 03
Multi-agent systems for real tasks
Guided mobile interaction, A2A-protocol multi-agent workflows, and modular robot communication, with evaluation protocol, observability, and deployment logic treated as first-class design goals.
Submission on Universal ReID under appearance variation
This project studies when structural cues should help person re-identification and when they should be prevented from distorting the main visual representation. The public layer keeps the problem setting, method boundary, and evaluation mindset visible, while title-level naming and implementation details are intentionally softened during review.
Under Review
PEARL: Verifiable curation and small-model learning for scientific reasoning graphs
Positioned as a complete framework organized around generation-verification decoupling, with training-free data construction layer and student SFT layer. Models propose reasoning graph candidates while code-side handles structural constraints, compilability, and source alignment.
In Writing
Survey in progress on appearance variation in person re-identification
This survey reorganizes how appearance variation should be grouped and evaluated, with an emphasis on clarifying the boundaries between robustness, generalization, and deployment evolution. The aim is not just a literature map, but a cleaner frame for future method design and experimental interpretation.
In Writing
Hands-on Work
Project Experience
I prefer building systems that are runnable, observable, and iteratively improvable.
A4R: Guided Smartphone AI Assistant
Designed and implemented an A4R interaction scheme for ambiguous mobile requests, built 150+ real interaction samples across 6 domains and 16 apps, and used Guide, Discern, and Validate collaboration to reach a 64.68% completion rate on China Unicom real-world tasks.
A2A Protocol Multi-Agent System Development
Built a Holos-based multi-agent system on A2A v0.3.0 with GuideAgent, DiscernAgent, and 8 functional agents; completed protocol adaptation, SSE streaming response, task-fingerprint deduplication, LLM/rule fallback, and observability debugging for travel-planning and email-collaboration demos.
Modular Robot System Development
Worked on software architecture for a modular intelligent hardware platform and a dual-bus power-communication system, implementing module coordination, bus arbitration, hot-plug, and bus-merging control logic while supporting three PCB iterations and 24V industrial-environment adaptation.
Xiaomi Quadruped Robot Intelligent Logistics Delivery System
Built a complete logistics delivery system on Xiaomi Quadruped Platform decomposed into departure, pickup, navigation, storage identification, unloading, return, and secondary delivery stages. Focused on state transitions, action timing coordination, and reducing failure propagation.
SmartPhone Battery: The Moment of Depletion
Built unified physical layer with dual-branch solvers: ECM-EKF for real-time SOC estimation and drift correction; PINN with identical dynamics for nonlinear fitting and factor attribution. Validated on 120,000+ time steps achieving 1.24% SOC RMSE and 8.7% TTE MAPE.
Background
Education and Recognition
I keep research, competitions, and systems implementation moving in parallel.
Education
Wuhan University
B.Eng. in Software Engineering, 2023.09 - 2027.06 (expected)
- GPA 3.74 / 4.0
- Research interests: visual representation under appearance variations, structured scientific reasoning with small models, multi-agent systems
- Core courses: Computer Vision, Frontier Technologies in AI, Computer Networks, Operating Systems, Software Engineering, SQA and Testing, Database Systems
Competitions
Selected Competition Achievements
- National University IoT Design Competition (Huawei Cup): National First Prize, Top 6
- China Youth Sci-Tech Innovation Challenge: National Second Prize
- National University Computer System Capability Competition (Xiaomi Cup): National Third Prize
- MCM 2026: Finalist
- MCM 2025: Honorable Mention
- CUMCM Hubei Division: Second Prize
Scholarships
Scholarships and University Honors
- Zheng Geru First-Class Scholarship, ranked 1st in grade defense and sole recipient
- Wuhan University First-Class Scholarship, twice
- Merit Student, twice
- Outstanding Student Cadre, Outstanding Youth League Branch Leader, Future College Outstanding Trainee
Skills and Service
Core Skills and Student Activities
Research work is supported by implementation depth, academic writing, and team coordination.
Core Skills
Engineering and Research Toolkit
- Python, C/C++, Java; Git and MATLAB
- Person Re-ID, SMPL body modeling, 3D Gaussian Splatting, experimental design, ablation analysis, academic writing
- Multi-agent design, orchestration, validation, observability, and application development with GPT, Claude, and Gemini APIs
- Embedded driver development and protocol design on RISC-V / ESP32 dual-architecture platforms
Student Activities
Representative Service Work
- Vice Minister of Publicity, Student Party Branch Secretaries Joint Council, Wuhan University (2024.09 - 2025.09)
- Supported undergraduate Party-building publicity work and contributed to more than 100 published articles
- Project Team Member, Second Classroom Center, University Youth League Committee (2023.10 - 2024.09)
- Participated in platform quality monitoring, issue investigation, and follow-through for platform fixes
Working Style
Where I tend to contribute best
Contact
Open to research collaboration, RA opportunities, and systems-oriented projects
If you are working on visual understanding, structured reasoning, or multi-agent systems, feel free to reach out.