Thang M. Pham

This is a likely match — the affiliation was inferred from OpenAlex, ORCID, and web sources but has not been fully confirmed. Treat with appropriate caution.

Researcher

University of Arkansas

Also affiliated: University of Engineering and Technology Lahore (2025); Auburn University (2021–2023)

Faculty Researcher

3 h-index 18 pubs 61 cited

Research Areas

OverviewAI-generated summary

Thang M. Pham's research focuses on developing and evaluating advanced techniques in machine learning and natural language processing. His work has explored real-time 3D mapping and queryable scene representation, as seen in his publication "Open-Fusion." Pham has also investigated the significance of sequential word order in natural language understanding and the creation of efficient language models for on-device applications, as demonstrated in "Out of Order" and "SlimLM." His research also addresses the explainability of text classifiers, particularly concerning counterfactual explanations generated by masked language models, as detailed in "Double Trouble." Pham has collaborated with Ngan Le, Winston Bounsavy, Minh Quan Tran, and Taisei Hanyu, all from the University of Arkansas at Fayetteville, on multiple publications. His h-index is 3 with 62 total citations across 19 publications.

Metrics

h-index: 3
Publications: 18
Citations: 61

Selected Publications

A2VIS: Amodal-Aware Approach to Video Instance Segmentation (2025)

Image and Vision Computing 2 citations DOI OpenAlex
A2vis: Amodal-Aware Approach to Video Instance Segmentation (2024)

SSRN Electronic Journal 1 citation DOI OpenAlex