Facebook AI Model Beats Google, Runs 5 Times Faster On GPUs

FAIR is advancing the state-of-the-art in artificial intelligence through fundamental and applied research in open collaboration with the community.

Photo: Getty Images

San Francisco: Facebook AI Research (FAIR) team recently developed a novel low-dimensional design space called ''RegNet'' that beats traditional available models like from Google and also it runs five times faster on GPUs. RegNet produces simple, fast and versatile networks and in experiments, it outperformed Google''s SOTA EfficientNet models, said the researchers in a paper titled ''Designing Network Design Spaces; published on pre-print repository ArXiv. The researchers aimed for "interpretability and to discover general design principles that describe networks that are simple, work well, and generalize across settings". The Facebook AI team convoyed controlled comparisons with EfficientNet with no training-time enhancements and under the same training setup. Launched in 2019, Google''s EfficientNet uses a combination of NAS and model scaling rules and represents the current SOTA. With comparable training settings and Flops, RegNet models outperformed EfficientNet models while being up to 5× faster on GPUs. Fairly than designing and developing individual networks, the team focused on designing actual network design spaces comprising huge and possibly infinite populations of model architectures. The Design space quality is examined using error empirical distribution function (EDF). While probing the RegNet design space also provided researchers other unexpected insights into network design. For instance, the researchers marked the depth of the best models is stable across compute regimes with an optimal depth of 20 blocks (60 layers). The paper further stated, "While it is common to see modern mobile networks employ inverted bottlenecks, researchers noticed that using inverted bottlenecks degrades performance. The best models do not use either a bottleneck or an inverted bottleneck". Facebook AI research team recently developed a tool that tricks the facial recognition system to wrongly identify a person in a video. The "de-identification" system, which also works in live videos, uses machine learning to change key facial features of a subject in a video. FAIR is advancing the state-of-the-art in artificial intelligence through fundamental and applied research in open collaboration with the community. The social networking giant formed the Facebook AI Research (FAIR) group in the year 2014 to elevate the state of the art of AI through open research for the benefit of all. Since then, FAIR has grown into an international research organization with labs in Menlo Park, New York, Paris, Montreal, Tel Aviv, Seattle, Pittsburgh, and London.