Adobe has good news for video editors, as it has announced plans to introduce new tools powered by generative AI (artificial intelligence) technology within the popular video editing platform Premiere Pro. These new tools will allow editors to add or remove an object from a video using a simple text prompt, powered by Firefly AI, including a new video generation model. It is expected that the new model will be available for Premiere Pro users starting in May 2024.
Currently, Adobe is also partnering with OpenAI and Runway. However, it will soon add more AI-backed video editing abilities to Premiere Pro in the coming days. This comprises an option to add a few more seconds to footage using the generative extend tool and the ability to generate a B-roll.
The new tool in Premiere Pro allows editors to quickly choose an object, track it, and remove an unwanted object from the frame with just a single input, like “quickly add set dressings such as photorealistic flowers or painting on a desk” or “change an actor’s wardrobe.” The object addition or removal generative extend feature can add frames to a video to make it longer, which is beneficial for tasks such as fine-tuning edits or designing a better transition between shots.
This is driven by both in-house Firefly models and third-party models such as OpenAI’s Sora, Pika, and the Runway large language model. Pika can extend shots, Sora, or Runway AI when generating a B-roll. Lastly, Adobe demonstrated a new text-to-video model that can produce entirely new footage within Premiere Pro by simply typing a text prompt or uploading a reference image.
Moreover, Adobe will soon introduce a new feature called content credentials, which makes it easy to identify if the content is original or created by AI using a watermark.
The new initiative by Adobe can help foster partnerships with third-party AI companies by explaining the third-party AI integrations in its video preview as an “early exploration” of what these might look like in the future.
The content credentials labels can be applied to the generated clips to identify which AI models have been used to generate them,” Adobe emphasised. In addition to that, Adobe has also demonstrated new audio workflows, including effect badges, interactive fade handles, essential sound badges with audio categories, and a redesigned waveform in the timeline.
Comments
0 comment