Software Engineer Computer Science

Gu, Yiwen Tel: 847-***-****

Email: ******@**.***

http://cs-people.bu.edu/yiweng/

EDUCATION

• Ph.D. candidate, Computer Science, Boston University (current)

• M.S., Computer Science, Boston University

• M.S., Biophysics, University of Wisconsin-Madison

• B.S., Biological Sciences, Fudan University, China WORKING EXPERIENCE

• Software Engineer Intern, SAIL, Hariri Institute for Computing, BU Feb 2019 – Aug 2019

• Backend Developer Intern, MIDA. Inc Jun 2018 – Aug 2018

• Bioinformatics Specialist, UW-Madison Jan 2016 – Aug 2017 RELEVANT RESEARCH PROJECTS

• ExeChecker

- Curated ExeCheck, a dataset of 10 rehab exercises for Parkinson’s patients, captured with MS Azure Kinect (RGB-D & 3D skeletons), featuring paired correct/incorrect movements.

- Built ExeChecker, a graph transformer model trained with triplet loss to distinguish correct from incorrect exercise forms.

- Proposed Joint-of-Attention (JoA) scores to measure model focus on relevant joints.

- Enabled interpretability by highlighting joints needing attention when incorrectness is detected. Keywords: Graph Transformers, Contrastive Learning, Human Motion Analysis, Explainability

• Rodent3D and OptiPose.

- Curated Rodent3D, a 200-min multi-modal (RGB, depth, thermal) dataset of rodent behavior (~4M frames).

- Developed OptiPose, a token-based model for 3D rodent pose estimation and temporal behavior tracking.

Keywords: Multiview Tracking, 3D Pose Estimation, Temporal Modeling, Animal Behavior

• ExerciseCheck.

- Re-engineered ExerciseCheck, a modular, user-friendly platform for remote physical therapy monitoring.

- Implemented real-time personalized feedback and redesigned the UI/UX for better patient engagement.

- Benchmarked four RGB-based pose estimators against Kinect (RGB-D), deployed and tested with Parkinson’s patients at home.

Keywords: Web application, HCI, Human Pose Estimation, Dynamic Time Warping

• Show, Write and Retrieve.

- Proposed a named entity-aware framework for news article generation and retrieval.

- Outperformed GPT, LLaMA, and CLIP on GoodNews, VisualNews, and WikiText datasets.

- Validated via user study on Amazon Mechanical Turk. Keywords: Entity-aware News Generation, Photos, Large Language Model. OTHER RESEARCH PROJECTS

• Movie-CLIP. – Keywords: Multi-label Classification, Language Augmentation, Shot Sampling

• Predicting Aphasia Recovery. – Keywords: fMRI, Model Evaluation. Feature Selection

• MASH Explorer. – Keywords: Software, Top-down Proteomics, C# TEACHING EXPERIENCE

• CS 640: Artificial Intelligence, Fall 2020, 2021, 2024

• CS 523: Deep Learning, Summer 2023

• CS 440: Artificial Intelligence, Spring 2025

• CS 111: Introduction to Computer Science, Spring 2020, Summer 2024

• Research in Science & Engineering Program (RISE), Summer 2020 AWARD AND CERTIFICATE

• Certificate of Completion, The Basic Drug Discovery Online Course.

• NSF Doctoral Consortium Travel Award

• Invited participant for the Grad Cohort Workshop of the CRA-W PROFESSIONAL SERVICE

• Seminar Chair (2023-2024) for the Artificial Intelligence Research (AIR) initiative at BU.

• Reviewer for Conferences/Journals :

– CVPR, WACV, MICCAI, CVPR/CV4Animals, ECCV/ACVR, PETRA

– IJCV, PAAA, PLOS ONE, Journal of Heliyon

PUBLICATIONS

• Y Gu, M Patel, M Betke. ExeChecker: Where Did I Go Wrong? Proceedings of the 12th International Workshop on Assistive Computer Vision and Robotics (ACVR/ECCV Workshops 2024)

• Z Zhang, Y Gu, B Plummer, X Miao, J Liu, H Wang. Movie genre classification by language augmentation and shot sampling. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2024)

• Z Zhang, Y Gu, B Plummer. Show, Write, and Retrieve: Entity-aware Article Generation and Retrieval. Findings of the Association for Computational Linguistics: EMNLP 2023

• M Patel, L Carstensen, Y Gu, M Hasselmo, M Betke. BU-CVKit: Extendable Computer Vision Framework for Species Independent Tracking and Analysis. arXiv:2306.04736 (2023).

• M Patel*, Y Gu*, L Carstensen, M Hasselmo, M Betke. Animal Pose Tracking: 3D Multimodal Dataset and Token-based Pose Optimization[1]. International Journal of Computer Vision,1-17. (IJCV 2022).

*: Co-first author

[1] This work was also accepted and presented on CV4Animals workshop (2022)

• Z Zhang, Y Gu, B Plummer, X Miao, J Liu, H Wang. Effectively leveraging Multi-modal Features for Movie Genre Classification. arXiv:2203.13281. (2022)

• A Billot, S Lai, M Varkanitsa, E Braun, M Bahrani, Y Gu, J Bighetto, B Rapp, T Parrish, D Caplan, C Thompson. WFNR–Communicative disorders/aphasia, communication. SOFMER–COVID-19.:18.

• Z Zhang, Y Gu, B Plummer. Show and Write: Entity-aware News Generation with Image Information. arXiv:2112.05917. (2021)

• Y Gu, M Bahrani, A Billot, S Lai, EJ Braun, M Varkanitsa, J Bighetto, B Rapp, TB Parrish, D Caplan, CK Thompson, S Kiran, M Betke. A machine learning approach for predicting post-stroke aphasia recovery: a pilot study. Proceedings of the 13th ACM International Conference on PErvasive Technologies Related to Assistive Environments. (2020).

• S McIlwain, Z Wu, Y Gu, S Ramanathan, X Liu, R Sun, I Ong, Y Ge. MASH Explorer: A Universal Software Environment for Top-Down Proteomics, Journal of Proteome Research. (2020).

• Y Gu, S Pandit, E Saraee, T Nordahl, T Ellis, M Betke. Home-Based Physical Therapy with an Interactive Computer Vision System. Proceedings of the IEEE International Conference on Computer Vision Workshops.

(2019).

• E Saraee, Y Gu, S Pandit, S Tran, E Shandelman, S Singh, TJ Nordahl, T Ellis, M Betke. ExerciseCheck: data analytics for a remote monitoring and evaluation platform for home-based physical therapy. Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments. (2019).

• S Pandit, S Tran, Y Gu, E Saraee, F Jansen, S Singh, S Cao, A Sadeghi, E Shandelman, T Ellis, M Betke. ExerciseCheck: A scalable platform for remote physical therapy deployed as a hybrid desktop and web application. Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments. (2019).

Contact this candidate