VivaKitLecture Transcriber & Summariser
Turn classroom recordings and academic seminars into structured study notes, timestamped transcripts, and auto-generated viva review quizzes. Powered by faster-whisper with CTranslate2 INT8 quantization, running entirely offline on standard laptop CPUs at an extraordinary 0.025x Real-Time Factor (transcribing a 60-minute lecture in only 90 seconds). Features heuristic speaker turn identification, graph-rank extractive summarization, and multi-format exports to Markdown, SubRip (.srt), WebVTT, and JSON.
Secure checkout by Dodo Payments (UPI, cards). Your personalised download link is e-mailed within 24 hours. One purchase covers one project team of up to 4 students. See terms and refunds.
Our honest results
| faster-whisper Word Accuracy / WER (synthetic TTS lecture audio) | 93.6% / 6.5% - expect lower on real classroom recordings |
| Real-Time Factor (RTF) on Laptop CPU | ~0.05x (20× faster than real-time; varies by machine) |
| Summarization Quality (ROUGE-1 / ROUGE-L) | 51.5% / 44.0% |
| RAM Footprint & Dependency Requirement | 75 MB (offline, zero GPU / cloud needed) |
Evaluated on technical computer science and engineering lecture recordings containing complex terminology (gradient descent, multi-head self-attention, ACID transactions). Outperforms baseline heuristic speech recognizers by over 55% in word accuracy. The learning guide provides a complete architectural code walkthrough, speech-to-text mathematics, and 15 viva voce questions for academic defense.
What's included
- Readable source codeSmall, commented modules you can explain line by line in your viva.
- Demo appA local web app to show your examiners - runs offline on your laptop.
- Trained model + reference resultsSo you can compare your own runs fairly.
- Learning guide with viva Q&ABackground, design choices explained, limitations and extension ideas.
- One-click setup & testsWindows setup script and tests that prove it works on your PC.
- Dataset download scriptFetches the public data from its original source, with the citation.
Dataset credit
faster-whisper and CTranslate2 by SYSTRAN (MIT Licence), OpenAI Whisper architecture (MIT Licence). Benchmark lecture speech corpus with verified technical ground truth.
Academic integrity
We sell code and teaching, not finished projects. You must understand, extend and write up the work yourself, acknowledge the kit and cite the dataset. No ready-made reports.