AI-generated imagery has become increasingly common like ‘Vertical Harry Potter Scenes’, with tools like Photoshop’s Generative Fill capable of filling images with AI-generated backgrounds. However, examples of AI-generated video are rarer. Recently, an ai artist shared a video demonstrating the use of the AI model StableDiffusion to fill in the tops and bottoms of shots from the Harry Potter film series, giving the impression that they were filmed in a vertical orientation. The results are surprisingly authentic-looking.
Generative Fill in Video:
The shared video reveals that only a small portion of the image is from the original movie footage, while the rest is generated by the AI model. The scenes range from hospital rooms to the iconic Hogwarts school hall, and Photoshop Beta effortlessly creates believable sets from the beloved film franchise. The level of detail and accuracy achieved by the AI model has impressed many viewers.
Potential Limitations:
While users admire the AI-generated footage, they also raise concerns about potential limitations. Some commenters note that the AI technique works well for static or locked camera shots, as well as those that can be tracked and stabilized. However, there may be a high variance in the model’s performance across different types of scenes. It is speculated that certain shots may be challenging for the AI model, resulting in inconsistent or botched results. The technique may also encounter problems when people cross the boundaries of the areas being filled, leading to flickering and inconsistencies.
Adobe Express With Firefly: Revolutionizing Content Creation and Collaboration
You can stay in touch with us for upcoming updates!
Conclusion:
The showcased use of AI model StableDiffusion in filling the tops and bottoms of Harry Potter film scenes to create a vertical orientation effect is highly impressive. By generating realistic sets, the AI model demonstrates its ability to enhance footage and deceive viewers into believing it was filmed in a different manner. However, potential limitations are acknowledged, particularly in scenes with movement or complex elements that may challenge the AI’s ability to generate consistent and accurate results. Nonetheless, this application showcases the continuous advancements in AI technology and its potential to revolutionize video editing and content creation.
