A degree of image identification at sub-human scales could be possible with more advanced clusters
This addresses the problem of achieving human-like visual recognition with limited resources for AI and computer vision researchers, but it is incremental as it builds on existing self-supervised techniques.
The research tackled whether self-supervised learning can achieve human-level visual comprehension with similar sensory input, by scaling both data volume and image quality, resulting in human-level object detection at sub-human scales using up to 200,000 images at 256 ppi.
The purpose of the research is to determine if currently available self-supervised learning techniques can accomplish human level comprehension of visual images using the same degree and amount of sensory input that people acquire from. Initial research on this topic solely considered data volume scaling. Here, we scale both the volume of data and the quality of the image. This scaling experiment is a self-supervised learning method that may be done without any outside financing. We find that scaling up data volume and picture resolution at the same time enables human-level item detection performance at sub-human sizes.We run a scaling experiment with vision transformers trained on up to 200000 images up to 256 ppi.