Google to Represent reCAPTCHA v3

As reported, reCAPTCHA v3 does not interfere with the user
30 October 2018   944

Google reported on a new API reCAPTCHA v3, which will help to protect websites from bots. The system evaluates the user's key actions on a scale from 0.0 to 1.0 (the lowest rating means a bot, and the highest - a person) and provides data to the owner for making a decision. Google experts recommend adding it to several pages of the site, as reCAPTCHA v3 does not interfere with the user.

Google developers used the “Action” method in reCAPTCHA v3. Detailed statistics of user interaction with the site are represented in the admin panel. The site owner independently decides on actions with respect to users. The grading system allows you to:

  • set the period of time after which the API will require the user to verify from the user using the phone or two-factor authentication;
  • combine actions based on them with manually specified parameters for users;
  • use machine learning to train AI, which distinguishes a bot from a person.

MelNet Algorithm to Simulate Person's Voice

It analyzes the spectrograms of the audio tracks of the usual TED Talks, notes the speech characteristics of the speaker and reproduces short replicas
11 June 2019   324

Facebook AI Research team has developed a MelNet algorithm that synthesizes speech with characteristics specific to a particular person. For example, it learned to imitate the voice of Bill Gates.

MelNet analyzes the spectrograms of the audio tracks of the usual TED Talks, notes the speech characteristics of the speaker and reproduces short replicas.

Just the length of the replicas limits capabilities of the algorithm. It reproduces short phrases very close to the original. However, the person's intonation changes when he speaks on different topics, with different moods, different pitches. The algorithm is not yet able to imitate this, therefore long sentences sound artificially.

MIT Technology Review notes that even such an algorithm can greatly affect services like voice bots. There just all communication is reduced to an exchange of short remarks.

A similar approach - analysis of speech spectrograms - was used by scientists from Google AI when working on the Translatotron algorithm. This AI is able to translate phrases from one language to another, preserving the peculiarities of the speaker's speech.