How is digital human technology changing film and television production?

I recently saw the OmniAvatar system developed by Zhejiang University and Alibaba, and it really opened my eyes! To be honest, as a movie and TV professional, I can't remember the last time I was so excited. This guy only needs a photo and a piece of audio to generate a full-body motion video, which is simply prying the foundation of the entire film and television production industry ah.

Do you guys remember Avatar back in the day when you had to wait days for a shot? The actors wore motion capture suits and performed in front of a blue screen, while the post-production team worked overtime to process the data. Now, OmmiAvatar technology, we may directly send a photo to the AI say "a sad solo dance", it can directly generate a complete digital human performance video. It's not just a matter of efficiency, it's a major shift in the way we create.

Digitizer technology puts post-production on a "diet"

I have to mention that OmniAvatar's accuracy in audio mouth synchronization is simply outrageous (and that's a compliment here). In the past, when we did post-production lip-syncing, we often had to rework a shot several times, and the voice actors had to record over and over again. Look at the current digital human technology, system-generated lip-synchronization is said to be able to reach a match of more than 90% with professional voice-overs, which is a boon for animation projects. If the animation director of Kung Fu Panda had seen this technology a few years earlier, I'm afraid he would have been moved to tears.

But what shocked me most was its ability to capture full-body motion. Traditional motion capture is not cheap: a decent set of motion capture equipment can cost millions of dollars, the venue conditions are demanding, and you have to hire professional motion capture actors. Now what? Maybe in the future, as long as the office to take a picture, say "add some sadness to his emotions", you can directly out of the movie. This cost difference, you do the math?

The double-edged sword of creative freedom

From a creative perspective, digital human technology has made it a whole lot less difficult for directors and writers to realize their ideas. Character designs and special performances that previously had to be abandoned because of budgetary or technical constraints may now be realized at the touch of a mouse. Think about it, if Gollum in The Lord of the Rings were to be done now, the production cycle could be cut in half.

But I don't think the technology is perfect. There is a concern that over-reliance on digital will diminish the realism of the performance. Currently, the system is not stable enough for long videos, and the "industrial feel" is still noticeable. Just as early CG effects always felt fake, digital human performances may need time to polish that subtle emotional expression.

The Ethical Dilemma of Digital Human Actors

Technology is good technology, but it can cause a lot of industry shock. Remember the Hollywood actors' strike? Part of the reason was concern about AI replacing actors. Now that the system can generate performances directly from photos, copyright issues, the scope of use of the right of portrait, the protection of the rights of actors ... these issues have to be rethought and redefined. The industry may soon see the emergence of new professions such as "digital human agents".

From the perspective of business model, it is an inevitable trend for film and television production to shift from asset-heavy to asset-light. Many small and medium-sized production companies may take this opportunity to bend the road and produce higher quality content at a lower cost. Of course, this also means that the industry will be more competitive. Whoever masters and skillfully utilizes digital human technology first will have a head start in this new era.