Voice Over Opportunities
Bhashini AI Solutions Pvt Ltd is looking for male and female voice over artists in Hindi,
Marathi*, Bengali, Telugu*, Kannada, Bhojpuri, Maithili, Magahi and Chhattisgarhi languages. The requirement is to record 55 hours of speech. This may take up to 110 hours of studio time depending on the fluency of the artist.
* Requirements for Marathi and Telugu languages are closed for both male and female artists.
The recording will happen at the BhaShiNi Studio located at:
Travel, Food and Accommodation:
Accommodation for the artist will be provided by the company for the duration of recording. One time travel expense of the artist to Bangalore and back will be reimbursed by the company. Food allowance of Rs.5000 will be provided.
Script to be recorded:
The script to be recorded will contain random sentences. The sentences usually are unrelated to each other. Entire script will be provided to the artist at the beginning. Artist should go through the script and make any necessary corrections to the script like punctuations, spellings, completeness etc.
Use of recorded speech:
The speech recorded as part of this project is intended to be used to train AI models that will be capable of generating speech in the voice of the artist. The collected speech may also be made available in public domain. The identity of the artist will not be revealed.
The artist will sign a legal agreement to give the following consent to use their voice:
1. The Artist is aware and agrees that recordings of their voice will be used by Bhashini and/or its clients to create and use a synthetic version of their voice.
2. The Artist is aware that a voice model made from their recordings can say things they didn't specifically record in the studio.
3. TTS use cases could include, but are not limited to:
● Developing assistive technologies for people with speech and visual disabilities (e.g., screen readers)
● Developing automatic dialogue systems where both speech recognition and TTS are key components. (e.g., voice-enabled queries and responses on weather, fertilizers, or crop price) etc.
● Virtual agents based on fictional personas (e.g., on-demand web searching, IoT control, or customer support provided by a company's branded character).
● Entertainment media for use in fictional content (e.g., movies, video games, tv, recorded music, or audio books)
● Accredited educational institutions or educational media (e.g., interactive lesson plans or guided museum tours)
● Assistive technology and real-time translation (e.g., ALS-afflicted individuals preserving their voices)
● Public service announcements using fictional personas (e.g., airport or train terminal announcements)
● TTS systems may be coupled with a computer-aided learning system to provide helpful tools for learning new languages by listening to the pronunciation of the words
● Designing different human–computer interactive systems (e.g., kiosks or automated tellers)
● TTS combined with speech recognition is useful for telecommunication and multimedia applications (e.g., synthetic speech used in games or talking toys)
Estimated duration for recording 55 hours of speech will be 30 working days. The actual recording time may vary depending on the fluency of the artist.
Recording and submission of voice samples:
Please see the video below for instructions on how to install and use the Vaak Sangraahak mobile app.