Members from the Google Brain and Google AI teams have announced they have open-sourced EfficientDet, an AI-powered tool that allows for state-of-the-art object detection with more efficient use of compute power.
The next generation version of EfficientNet, EfficientDet was first detailed in a paper published last fall by Google engineers Mingxing Tan, Ruoming Pang, and Quoc Le.
“Aiming at optimizing both accuracy and efficiency, we would like to develop a family of models that can meet a wide spectrum of resource constraints,” the paper reads.
According to the authors, while other methods of object detection sacrifice accuracy and are resource-intensive as they scale, EfficientDet uses a method that “uniformly scales the resolution, depth, and width for all backbone, feature network, and box/class prediction networks at the same time,” while consuming fewer computing resources.
“The large model sizes and expensive computation costs deter their deployment in many real-world applications such as robotics and self-driving cars where model size and latency are highly constrained,” the paper reads, going on to say that “[g]iven these real-world resource constraints, model efficiency becomes increasingly important for object detection.”