Bengaluru-based NeuralGarage Dubs War 2 in Telugu Using AI

Bengaluru-based NeuralGarage Dubs War 2 in Telugu Using AI

Analytics India Magazine (Siddharth Jindal)

Bengaluru-based AI startup NeuralGarage has converted the entire Hindi film War 2 into Telugu, making it appear as if the movie was originally shot in Telugu. The film, releasing on August 14, the company said, received a ‘straight’ film certificate from the censor board.

“The entire movie was shot in Hindi, including the scenes of Junior NTR. We lip-synced the full movie to Telugu so that it looks like it has been originally shot in Telugu,” Subhabrata Debnath, co-founder and CTO of NeuralGarage, told AIM.

VisualDub, the company’s proprietary video transformation model, alters actors’ facial movements using audio as a controller. NeuralGarage said it enables multi-lingual film distribution, dialogue replacement, and advertising applications without re-shoots while preserving the original performance.

The technology has previously been used by Yash Raj Films on War 2, Sun TV to dub Rajnikanth’s Coolie song Chikitu in Hindi and Telugu, and Dharma Productions for Kesari 2, making English-speaking actors appear to speak in Hindi. 

It was also used in Neeraj Pandey’s Special Ops to correct the actor’s facial movements for censored scenes. The company claims over 50 enterprise clients, including Amazon, Unilever, P&G, and Coca-Cola.

“VisualDub enables actors to speak to new markets without re-shoots, while keeping brand and face continuity across regions and unlocking roles beyond native language,” Debnath said.

NeuralGarage emphasised that achieving theatrical-quality visual dubbing requires addressing frame quality, colour and data integrity, and editorial interoperability.

“Done right, visual dubbing doesn’t just translate dialogue—it extends an actor’s market and deepens connection with local audiences at theatrical quality,” Debnath added.

The post Bengaluru-based NeuralGarage Dubs War 2 in Telugu Using AI appeared first on Analytics India Magazine.

Generated by RSStT. The copyright belongs to the original author.

Source

Report Page