ViBR: Automated Bug Replay from Video-based Reports using Vision-Language Models DOI: 10.1145/3808152 Journal: PACMSE Volume: 3 FSE FSE145 7 fse26mainb-p692-p CCS: Software and its engineering Software testing and debugging
Sidong Feng, Dingbang Wang, Nikola Tomic, Tingting Yu, Aldeida Aleti, Chunyang Chen · 2026
Read ViBR: Automated Bug Replay from Video-based Reports using Vision-Language Models DOI: 10.1145/3808152 Journal: PACMSE Volume: 3 FSE FSE145 7 fse26mainb-p692-p CCS: Software and its engineering Software testing and debugging by Sidong Feng, Dingbang Wang, Nikola Tomic, Tingting Yu, Aldeida Aleti, Chunyang Chen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-19905.txt.
Description
Feng and five colleagues built ViBR, a tool that replays software bugs straight from screen-capture videos. Such videos are a popular bug-report format, but hard to replay by machine. ViBR cuts actions apart with CLIP embedding similarity, then checks interface states and guides replay with vision-language models. It reproduced 72 percent of test recordings, well ahead of leading baselines and ablated variants.
Catalog record
| CREATOR | Sidong Feng, Dingbang Wang, Nikola Tomic, Tingting Yu, Aldeida Aleti, Chunyang Chen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69003 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.19905