· 2020
EfficientDet: Scalable and Efficient Object Detection
- 8,554citations
- 2020year
Short summary
EfficientDet detectors achieve state-of-the-art COCO AP (52.2) with 4x-9x fewer parameters and 13x-42x fewer FLOPs than prior detectors, by using a weighted bi-directional feature pyramid network (BiFPN) and a compound scaling method.
AI-generated from the title and abstract; the full text is not read.
Key points
- Introduced a weighted bi-directional feature pyramid network (BiFPN) for efficient multi-scale feature fusion.
- Developed a compound scaling method to uniformly scale network width, depth, and resolution.
- EfficientDet-D7 achieved 52.2 AP on COCO test-dev with 52M parameters and 325B FLOPs.
- Achieved 4x-9x fewer parameters and 13x-42x fewer FLOPs than previous state-of-the-art detectors.
AI-generated from the title and abstract; the full text is not read.
Abstract
Model efficiency has become increasingly important in computer vision. In this paper, we systematically study neural network architecture design choices for object detection and propose several key optimizations to improve efficiency. First, we propose a weighted bi-directional feature pyramid network (BiFPN), which allows easy and fast multi-scale feature fusion; Second, we propose a compound scaling method that uniformly scales the resolution, depth, and width for all backbone, feature network, and box/class prediction networks at the same time. Based on these optimizations and EfficientNet backbones, we have developed a new family of object detectors, called EfficientDet, which consistently achieve much better efficiency than prior art across a wide spectrum of resource constraints. In particular, with single-model and single-scale, our EfficientDet-D7 achieves state-of-the-art 52.2 AP on COCO test-dev with 52M parameters and 325B FLOPs1, being 4x - 9x smaller and using 13x - 42x fewer FLOPs than previous detector. Code is available at https://github.com/google/ automl/tree/master/efficientdet.
The authors' abstract, as published at the source., 2020 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionComputer Science