Lecture Transcription/Summarization (Project)
Developed an automated pipeline to transcribe and summarize audio files by combining Whisper for speech-to-text and T5 for abstractive summarization. The system produces summarized text outputs derived from spoken lecture content. This involved AI training/data processing style work around transcription-to-summary generation. • Transcribed audio with Whisper. • Summarized transcribed text with T5 (abstractive summarization). • Built an end-to-end processing workflow for non-technical users. • Provided a Flask interface to upload and generate summaries.