The cooperation between Meitu Imaging Research Institute and Beijing Jiaotong University has yielded fruitful results. The MEMatte ultra-high-resolution matting technology they jointly developed was successfully selected for the AAAI2025 conference. This technology breaks through the computing resource limitations of traditional matting methods, enabling it to achieve fine matting of high-definition images on devices with limited resources, bringing innovation to fields such as video production, virtual reality, and augmented reality. MEMatte technology is efficient and memory-friendly. At the same time, the research team has also open sourced the ultra-high-resolution natural image matting data set UHR-395, providing valuable resources to promote technological development in this field.
Recently, Meitu Imaging Research Institute (MT Lab) and Beijing Jiaotong University jointly proposed an ultra-high-resolution matting technology called MEMatte (Memory Efficient Matting), which was successfully selected for AAAI2025, the top conference in the field of artificial intelligence. The biggest highlight of MEMatte technology is that it is a memory-friendly natural image matting framework that can effectively reduce the computational overhead of the model. This innovation enables detailed matting of HD images in memory-constrained environments such as commodity graphics cards and edge devices.
With the continuous development of image processing technology, cutout technology has been widely used in many fields, such as video production, virtual reality and augmented reality. However, traditional matting methods usually require a large amount of computing resources, which makes them difficult to implement in some resource-limited scenarios. MEMatte was developed to address this problem, improving processing efficiency while maintaining the quality of high-resolution images.
In addition, the research team also open sourced an ultra-high-resolution natural image matting dataset called UHR-395 (Ultra High Resolution dataset). The launch of this data set will provide valuable resources for the training and evaluation of high-resolution models and promote the further development of related technologies. Through open source, the research team hopes to attract more researchers and developers to participate in this field and jointly promote technological progress.
Highlight:
1. Meitu Imaging Research Institute and Beijing Jiaotong University jointly developed MEMatte technology, which has been selected for the AAAI2025 conference.
2. MEMatte technology is memory-friendly, can effectively reduce computing overhead, and is suitable for devices with limited resources.
3. The open source ultra-high-resolution matting data set UHR-395 facilitates the training and evaluation of high-resolution models.
The emergence of MEMatte technology and the open source of the UHR-395 data set mark a new milestone in high-resolution matting technology, providing broader possibilities for the application of future image processing technology. I believe that with the joint efforts of more researchers and developers, this technology will bring more surprising applications.