I have previously introduced Domo AI, an AI video style conversion tool that can convert real-life videos into various virtual styles. It is worth noting that this tool is mainly targeted at the overseas market, thanks to its successful marketing on social media platforms. In just three months, the number of registered users has exceeded 100,000. However, perhaps due to the rapid growth of users, it has led to a sharp increase in computational costs. A few days ago, Domo AI temporarily suspended free trials for new users. Today, I discovered a perfect alternative to Domo AI - GoEnhance, which offers limited free trials and currently supports nine types of video style conversions.
GoEnhance preserves the details of the original video better, including the texture and patterns of clothing, accessories and hair of the characters. It retains them more completely, and the coherence of the images is also better. It is difficult to see elements disappearing or jumping in the images. Interested users can try it out for themselves. New users can generate three 3-second videos or two 5-second videos for free, and it supports web access (which has become an advantage). Unlike Domo AI, which only supports Discord, GoEnhance currently only supports web access and requires a Google account to log in.
In addition, GoEnhance has a nice detail in terms of user experience - the generated videos come with the original audio. In the scenes where AI video conversion is used, such as dance videos, there is often background music. GoEnhance captures the original audio track of the video and combines it with the converted video before providing it to the user. The downside is that the generation speed is slightly slow, which may be related to computational resources or the need to handle more details technically. Currently, there are still few users, but as the user base grows, this may become a significant bottleneck.
In addition to supporting Video2Video video style conversion, GoEnhance also has an image upscale function. Considering its name and web style, I suddenly realized that the development team initially wanted to create an image enhancement tool similar to KREA AI or Magnific AI. Video2Video video style conversion has more entertainment value and can be used to adapt dance videos on platforms like TikTok, reaching a wider audience. However, it is not a strong demand and copyright issues still exist. On the other hand, Image Upscale has more productivity attributes and can be used for image enlargement or restoring old photos, but it has a narrower audience and is more geared towards professional users who have higher performance requirements. If there is not enough differentiation from Magnific AI, it will be difficult to survive.
These two issues mentioned above may exist in many AI painting or video applications, including GoEnhance, which is learning from Domo AI. The high computational cost and difficulty in converting free users into paying customers make it necessary to first build up the user base through unique features and marketing, and then secure funding. There are not many choices available at the moment. How many startup companies can be as successful as Pika and Magnific AI?
What are your thoughts on GoEnhance or AI applications? Feel free to share and repost.