EfficientDet: Scalable and Efficient Object Detection EfficientDet is a new family of object detectors that achieve significantly better efficiency than previous methods. The paper introduces a weighted bidirectional feature pyramid network (BiFPN) for efficient multi-scale feature fusion and a compound scaling method that uniformly scales resolution, depth, and width for all backbone, feature network, and box/class prediction networks. These optimizations, combined with better backbones, enable EfficientDet to consistently achieve much better efficiency across a wide spectrum of resource constraints. EfficientDet-D7 achieves state-of-the-art 55.1 AP on COCO test-dev with 77M parameters and 410B FLOPs, being 4x-9x smaller and using 13x-42x fewer FLOPs than previous detectors. EfficientDet is also up to 4x-11x faster on GPU/CPU than previous detectors. The model is also effective for semantic segmentation, achieving 81.74% mIOU accuracy with 18B FLOPs on Pascal VOC 2012. The paper also discusses the design of EfficientDet, including the BiFPN architecture and compound scaling method, and presents results on COCO and Pascal VOC datasets. The paper concludes that EfficientDet achieves better accuracy and efficiency than previous detectors across a wide range of resource constraints.EfficientDet: Scalable and Efficient Object Detection EfficientDet is a new family of object detectors that achieve significantly better efficiency than previous methods. The paper introduces a weighted bidirectional feature pyramid network (BiFPN) for efficient multi-scale feature fusion and a compound scaling method that uniformly scales resolution, depth, and width for all backbone, feature network, and box/class prediction networks. These optimizations, combined with better backbones, enable EfficientDet to consistently achieve much better efficiency across a wide spectrum of resource constraints. EfficientDet-D7 achieves state-of-the-art 55.1 AP on COCO test-dev with 77M parameters and 410B FLOPs, being 4x-9x smaller and using 13x-42x fewer FLOPs than previous detectors. EfficientDet is also up to 4x-11x faster on GPU/CPU than previous detectors. The model is also effective for semantic segmentation, achieving 81.74% mIOU accuracy with 18B FLOPs on Pascal VOC 2012. The paper also discusses the design of EfficientDet, including the BiFPN architecture and compound scaling method, and presents results on COCO and Pascal VOC datasets. The paper concludes that EfficientDet achieves better accuracy and efficiency than previous detectors across a wide range of resource constraints.