Summary of Transcription and Translation Of Videos Using Fine-tuned Xlsr Wav2vec2 on Custom Dataset and Mbart, by Aniket Tathe et al.
Transcription and translation of videos using fine-tuned XLSR Wav2Vec2 on custom dataset and mBART
by Aniket Tathe, Anand Kamble, Suyash Kumbharkar, Atharva Bhandare, Anirban C. Mitra
First submitted to arxiv on: 1 Mar 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Sound (cs.SD); Audio and Speech Processing (eess.AS)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper presents a method to train automatic speech recognition (ASR) models for personalized voices using minimal data. The approach involves creating a custom audio dataset from a YouTube video and fine-tuning a cross-lingual self-supervised representation (XLSR) Wav2Vec2 model on this dataset. The system is designed with a web-based graphical user interface (GUI) that efficiently transcribes and translates input Hindi videos, aligning the translated text with the video timeline. This accessible solution enables multilingual video content transcription and translation for personalized voice applications. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper makes it possible to train an ASR model for personalized voices using just 14 minutes of custom audio from a YouTube video. The system uses a technique called Retrieval-Based Voice Conversion (RVC) to create a special dataset. Then, a type of AI model called Cross-lingual Self-supervised Representations (XLSR) Wav2Vec2 is adjusted to work well with this data. The system also has a user-friendly interface that can translate and transcribe videos in Hindi. This makes it easier for people to understand videos in different languages. |
Keywords
* Artificial intelligence * Fine tuning * Self supervised * Translation
