Windows Desktop Application
Advanced Video QA System
Desktop AI video question-answering application with multimodal audio-visual indexing, exact timestamp citations, and a standalone Windows installer.
Problem
Finding specific moments or verifying statements across hours of video lectures, webinars, or surveillance footage requires tedious manual scrubbing.
Implementation and contribution
Video ingestion, transcript retrieval and timestamp-based question answering are described in the existing project documentation and resume. The portfolio includes desktop workflow screenshots and a Windows installer.
Architecture
Python Multiprocessing + Whisper Audio Transcription + Vision LLM Provider + InnoSetup Packaging
Trade-offs and limitations
Transcription and model responses may be inaccurate. Model-provider features can require an API connection; a desktop package does not imply all inference is offline.
Evidence and availability
Source repository and release download are available. Installer execution, inference quality and timing were not independently tested in this portfolio review.
Technologies
- Python
- PyQt / Desktop UX
- OpenCV
- Whisper
- Gemini Vision
- InnoSetup
- Windows Desktop
Workflow screenshots
Descriptions are grounded in the existing portfolio and available project material. Unsupported benchmarks are not presented as verified results.
Discuss this project