Bengaluru, June 27: The AI-powered chatbots are on the rise with the advancements in the field of artificial intelligence. Many forms of generative AI tools are being released by various tech majors, but the most popular among these are the text based generative AI, that can process and generate text with lucid language that is very similar to human conversation or human-created text. It is but natural that tech giant Google will not be behind in this field.

The advancements continue in large language models (LLM), as there’s a huge push towards their development recently, especially after the mega success of Microsoft backed OpenAI’s ChatGPT. Google has released a slew of various models or tools that are powered by AI. LLM utilizes an artificial neural network that works similar to a human brain’s section and processes and generates language. This artificial neural network is then trained extensively on a very large amount of text data using self-supervised learning. Google Announces ‘Perspectives’ Filter on Search; Know About Latest Update.

What is Google's AudioPaLM & How It Works:

Google’s Bard, the rival to ChatGPT is based on such a large language model, while it is also used partly in many other recent tools that Google has released. Very recently, the tech giant has introduced a new language model called AudioPaLM, which is multifaceted in relation to text and can listen, speak, and translate text in a natural human-like manner.

AudioPaLM is a multimodal architecture, which combines PaLM-2 and AudioLM and cannot only produce text but also speech, and can be used for speech recognition and translations of human speech. Byju’s CEO Raveendran Admits Mistakes, Assures Quick Resolution, CFO Ajay Goel Offers Deadlines for Closing Audits at Shareholder Meet: Reports.

PaLM-2 is a text-based language model which excels in text based language comprehension, while AudioLM is great at retaining paralinguistic data such as speaker identity and tone.

AudioPaLM offers greater usability and better prowess in various language related functions by combining these two models. Hence, it can speech-to-text translations for several languages, even for speech/language combinations that the tool was not trained for. This model can be effectively used in real-world scenarios in terms of real-time multilingual communication.

AudioPaLM can also effectively capture and reproduce different voices in different languages and is superb at speech translations, as has been demonstrated by Google’s researchers.

(The above story first appeared on LatestLY on Jun 27, 2023 07:30 AM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).