Kaldi Reviews and Details

Is Kaldi good?

Social recommendations and mentions

• Cloud Solutions vs. On-Premise Speech Recognition Systems

  Use of Open-Source Solutions and Customizable Models. On-premise systems, such as Lingvanex and Kaldi, provide tools to develop speech recognition models from scratch or based on open-source libraries. Unlike cloud services, where developers are limited to pre-built models, on-premise solutions allow you to create a system that fully matches the specifics of the task. For example, models can be trained on specific... - Source: dev.to / 7 months ago

• Amazon plans to charge for Alexa in June–unless internal conflict delays revamp

  Yeah, whisper is the closest thing we have, but even it requires more processing power than is present in most of these edge devices in order to feel smooth. I've started a voice interface project on a Raspberry Pi 4, and it takes about 3 seconds to produce a result. That's impressive, but not fast enough for Alexa. From what I gather a Pi 5 can do it in 1.5 seconds, which is closer, so I suspect it's only a... - Source: Hacker News / over 1 year ago

• Steve's Explanation of the Viterbi Algorithm

  You can study CTC in isolation, ignoring all the HMM background. That is how CTC was also originally introduced, by mostly ignoring any of the existing HMM literature. So e.g. Look at the original CTC paper. But I think the distill.pub article (https://distill.pub/2017/ctc/) is also good. For studying HMMs, any speech recognition lecture should cover that. We teach that at RWTH Aachen University but I don't think... - Source: Hacker News / almost 2 years ago

• Best text to speech softwares

  I also tried Kaldi but the build process was too much for my tiny brain; I've also heard good things about vosk but didn't try that. Source: about 2 years ago

• The Advantages and disadvantages of In-House Speech Acknowledgment

  Frameworks as well as toolkits like Kaldi were at first promoted by the research study area, yet nowadays used by both scientists and also market experts, reduced the access obstacle in the advancement of automatic speech recognition systems. Nonetheless, cutting edge methods need big speech data readies to achieve a usable system. Source: over 2 years ago

• Machine Learning with Unix Pipes

  If you interested in unix-like software design and not yet familiar with kaldi toolkit, you definitely need to check it https://kaldi-asr.org It extended Unix design with archives, control lists and matrices and enabled really flexible unix-like processing. For example, recognition of a dataset looks like this: extract-wav scp:list.scp ark:- | compute-mfcc-feats ark:- ark:- | lattice-decoder-faster final.mdl... - Source: Hacker News / over 2 years ago

• Lexicap: Lex Fridman Podcast Whisper Captions by Andrej Karpathy

  No, speaker diarization is not part of Whisper. There are open source projects - such as Kaldi [1], but it's hard to get them running if you are not an area expert. [1] https://kaldi-asr.org/. - Source: Hacker News / almost 3 years ago

• Is there a way to integrate a raspberry pi with a keyboard to do speech to text?

  State-of-the-art ASR, like what you get on smartphones, has unfortunately high resource requirements. Some recent smartphone models are able to run ASR on-device, but more typically, ASR is done by sending audio to a web service. Check out the (currently experimental) Web SpeechRecognition API in a Chrome browser. Here is a demo of the API in action. For something open source, check out Kaldi ASR. Source: almost 3 years ago

• 5 Best Open Source Libraries and APIs for Speaker Diarization

  Kaldi ASR is a well-known open source Speech Recognition platform. To use its Speaker Diarization library, you’ll need to either download their PLDA backend or pre-trained X-Vectors, or train your own models. - Source: dev.to / over 3 years ago

• I made a tutorial on how to do Speech Recognition with Python and Kaldi!

  Kaldi is a really powerful toolkit for ASR and related NLP tasks, but I've found that the learning curve is a bit steep. I made a tutorial that you can find here that takes you through installation and transcription using pre-trained models, but the cool part is that you can decide how advanced you want it to be! Source: over 3 years ago

• Help picking a good speech recognition library

  Https://kaldi-asr.org/ (best out of the box accuracy but it is a complicated toolkit and not beginner friendly). Source: over 3 years ago

• Ask HN: What problem are you close to solving and how can we help?

  I worked on this for a couple years during a previous startup attempt. I designed a custom STT model via Kaldi [0] and hosted it using a modified version of this server [1]. I deployed it to a 4GB EC2 instance with configurable docker layers (one for core utils, one for speech utils, one for the model) so we could spin up as many servers as we needed for each language. I would recommend the WebRTC or Gstreamer... - Source: Hacker News / almost 4 years ago

• [D] ASR/Automatic Speech Recognition toolkit that provides precise word-level timing data? (eg, where in the audio stream a word starts and ends?)

  It sounds like you could use forced alignment, which can be done through Kaldi or the Montreal Forced Aligner, which uses Kaldi for backend ASR. Full disclosure, I'm the primary maintainer for MFA, but it should fit your use case. Source: almost 4 years ago