Ai text-to-video is developing fast - very fast! In the space of just a year of this newsletter, a plethora of tools have either become available or have been announced. It seems like a good time to kick the tyres of one of the Ai Video tools available, and have a little fun.
Today I’m taking a look at Haiper AI.
Haiper TL;DR
“Founded by alumni from Google DeepMind, TikTok and top research labs from academia, Haiper brings both cutting edge machine learning and a belief that creativity should be fun, surprising and shareable.”
Features
Generative AI for Video: Haiper.ai provides tools for generating video content, emphasizing the creative unleashing of mankind's creativity.
Drag and Drop Images: Users can drag and drop their images to bring them to life.
Video Repainting: This feature allows users to modify colors, textures, and elements in their videos easily, transforming the visual content significantly.
Text-to-Video Creation: Haiper.ai offers a simple process for creating videos from text prompts, making it easy to generate content quickly.
Benefits
Haiper.ai's technology is designed to augment human creativity, providing tools that make it easier to create engaging and innovative video content.
The platform's user-friendly features, like drag-and-drop image animation and video repainting, make it accessible for users with varying levels of technical expertise.
With the ability to create videos from text in minutes, Haiper.ai facilitates rapid content generation, allowing for more efficient workflow.
… and at the moment, it’s free to use 😁
The Platform
I decided to use Haiper to create a tongue-in-cheek spoof SciFi movie trailer. My DNA drives me to find the path of least resistance, so I decided to use an iMovie template for the trailer. If you aren’t familiar with iMovie, there are a number of templates to choose from. Each template has a shot list of clips needed for it (stills or video), and it includes music and editable credits.
For my blockbuster, I needed about 30 clips… enter Haiper.
The Haiper platform is really easy to use. There are a number of options to choose from: Video from text, Animate a still image, Change the style of an existing video.
I opted for create a video from text. A prompt box pops up, and as is normal now, enter a description of the video you want the Ai to create. There’s a settings box where you can increase or decrease motion, and alter the length of the clip (currently 2 seconds or 4 seconds). There’s also an option to add a seed. You can either just go with your text prompt, or add a preset to it such as ‘Steampunk,’ ‘Old Film,’ ‘Arcane,’ etc. I opted for ‘Cyberpunk’.
I set Haiper to work creating videos. Each video took around 2 minutes or so to appear. The first thing to note is that the Ai is a tad flakey at the moment. Some videos were spot on as per the prompt, others were frustratingly not. I guess that’s just the state of AI, right? As it was the first time I had used Haiper, it gets a gimme because it may be better if I played with it more.
It took maybe a couple of hours to put the shots together. I used Midjourney to create still images for a couple of the more frustrating shots, and then fed them into Haiper’s ‘Animate’ tool.
Then it was simply a case of dropping them into iMovie, changing the text, and exporting.
Here’s the finished product …
Conclusions
Haiper is a great tool, and maybe with a little more effort on my part, I can coax it to produce better results every time. It still needs lots of work, and competition from the likes of Stable Diffusion and soon, Midjourney, is fierce. But considering that this was my first time with the tool, I think it’s pretty darn good and definitely one to watch!