Environment perception on urban roads is one of the most important tasks in Automatic Driving System, because object detection is necessary in identifying and localizing cars and pedestrians. However, the accuracy and inference speed are still far away from satisfying, and the problem of missed detection occurs frequently. To address the problems, this paper first uniformly scales all dimensions of depth, width, and resolution in feature extraction module, and then fuses the semantic information in bidirectional cross-scale connections, which extract more significant features and detect objects of various scales. To decrease the undetected rate of small objects, this paper further introduces an IoU-based center localization confidence, which predicts centerness and the distance between the center of predicted box and the center of the ground-truth box, so that the predicted point approaches the center of an object. The experimental results on Cityscapes show that the proposed detector brings an improvement of 6.5% Bbox mAP and reaches an inference speed of 36.6 FPS. Compared with the other widely-used detectors such as Cascade R-CNN, FCOS, and YOLOX, it has obvious advantages in both precision and real-time performance. With additional ticks, the detection precision can be further improved by 4.4%. The proposed detector can accurately and quickly identify and localize objects, which is beneficial for the safety of self-driving cars.
LiJMeiXProkhorovD, et al. Deep neural network for structural prediction and lane detection in traffic scene. IEEE Trans Neural Netw Learn Syst2017; 28(3): 690–703.
2.
LvZHuangXLiangY, et al. An improved efficient model for structure-aware lane detection of unmanned vehicles. Proc IMechE, Part D: J Automobile Engineering2021; 235(9): 2496–2508.
3.
GengKZouWYinG, et al. Low-observable targets detection for autonomous vehicles based on dual-modal sensor fusion with deep learning approach. Proc IMechE, Part D: J Automobile Engineering2019; 233(9): 2270–2283.
4.
YangZLiJLiH. Real-time pedestrian and vehicle detection for autonomous driving 2018. In: IEEE intelligent vehicles symposium (IV), 2018, pp.179–184. New York: IEEE.
5.
LinCShiYZhangJ, et al. An anchor-free detector and R-CNN integrated neural network architecture for environmental perception of urban roads. Proc IMechE, Part D: J Automobile Engineering2021; 235(12): 2964–2973.
6.
ViolaPJonesMJ. Robust real-time face detection. Int J Comput Vis2004; 57(2): 137–154.
7.
DollarPAppelRBelongieS, et al. Fast feature pyramids for object detection. IEEE Trans Pattern Anal Mach Intell2014; 36(8): 1532–1545.
8.
JiaoXYangDJiangK, et al. Real-time lane detection and tracking for autonomous vehicle applications. Proc IMechE, Part D: J Automobile Engineering2019; 233(9): 2301–2311.
9.
ChenJAiYQianY, et al. A novel Siamese attention network for visual object tracking of autonomous vehicles. Proc IMechE, Part D: J Automobile Engineering2021; 235(10-11): 2764–2775.
10.
Zhong-QiuZPengZShou-TaoX, et al. Object detection with deep learning: a review. Computer Vision and Pattern Recognition. arXiv preprint arXiv: 1807.05511, 2018.
11.
GirshickR. Fast R-CNN. In: Proceedings of the IEEE international conference on computer vision, 2015, pp.1440–1448. New York: IEEE.
12.
RenSHeKGirshickR, et al. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE transactions on pattern analysis and machine intelligence2017; 39(6): 1137–1149. New York: IEEE.
13.
KhosravianAAmirkhaniAMasih-TehraniM. Enhancing the robustness of the convolutional neural networks for traffic sign detection. Proc IMechE, Part D: J Automobile Engineering2022; 236: 1849–1861.
14.
LiuWAnguelovDErhanD, et al. SSD: Single shot multibox detector. In: LeibeBMatasJSebeN, et al. (eds) European conference on computer vision. Cham: Springer, 2016, pp.21–37.
15.
RedmonJFarhadiA. Yolov3: an incremental improvement. arXiv preprint arXiv: 1804.02767, 2018.
16.
LinTYGoyalPGirshickR, et al. Focal loss for dense object detection. In: Proceedings of the IEEE international conference on computer vision, 2017, pp.2980–2988. New York: IEEE.
17.
HeKZhangXRenS, et al. Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp.770–778. New York: IEEE.
18.
LinTYDollárPGirshickR, et al. Feature pyramid networks for object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp.2117–2125. New York: IEEE.
19.
TianZShenCChenH, et al. FCOS: fully convolutional one-stage object detection. In: Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp.9627–9636. New York: IEEE.
20.
ZhangSChiCYaoY, et al. Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp.9759–9768. New York: IEEE.
21.
NeubeckAVan GoolL. Efficient non-maximum suppression. In: 18th international conference on pattern recognition (ICPR’06), 2006, vol. 6, pp.850–855. New York: IEEE.
22.
LiXLvCWangW, et al. Generalized focal loss: towards efficient representation learning for dense object detection. In: IEEE transactions on pattern analysis and machine intelligence, 2022, pp.1–14. New York: IEEE.
23.
HuangZHuangLGongY, et al. Mask scoring R-CNN. In: Proceedings of the IEEE conference on computer vision and pattern recognition, 2019, pp.6402–6411. New York: IEEE.
24.
HeKGkioxariGDollárP, et al. Mask R-CNN. In: Proceedings of the IEEE international conference on computer vision, 2017, pp.2961–2969. New York: IEEE.
25.
JiangBLuoRMaoJ, et al. Acquisition of localization confidence for accurate object detection. In: Proceedings of the European conference on computer vision (ECCV), 2018, pp.784–799. Springer.
26.
RezatofighiHTsoiNGwakJY, et al. Generalized intersection over union: a metric and a loss for bounding box regression. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp.658–666. New York: IEEE.
27.
ZhengZWangPLiuW, et al. Distance-IoU loss: faster and better learning for bounding box regression. Proc AAAI Conf Artif Intell2020; 34: 12993–13000.
28.
TanMPangRLeQV. Efficientdet: scalable and efficient object detection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp.10781–10790. New York: IEEE.
HuangYChengYBapnaA, et al. Gpipe: efficient training of giant neural networks using pipeline parallelism. In: WallachHLarochelleHBeygelzimerA, et al. (eds) Advances in neural information processing systems 32. Curran Associates, Inc, Vancouver 2019, pp.103–112.
31.
TanMLeQ. Efficientnet: rethinking model scaling for convolutional neural networks. In: International conference on machine learning, PMLR, Association for Computing Machinery(ACM) 2019, pp.6105–6114.
32.
HowardA GZhuMChenB, et al. Mobilenets: efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv: 1704.04861, 2017.
33.
CordtsMOmranMRamosS, et al. The Cityscapes dataset for semantic urban scene understanding. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2016, pp.3213–3223. New York: IEEE.
34.
GeigerALenzPStillerC, et al. Vision meets robotics: the KITTI dataset. Int J Rob Res2013; 32(11): 1231–1237.
35.
YangZLiuSHuH, et al. Reppoints: point set representation for object detection. In: Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp.9657–9666. New York: IEEE.
36.
CaiZVasconcelosN. Cascade R-CNN: delving into high quality object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp.6154–6162. New York: IEEE.
37.
TanMLeQ. Efficientnetv2: smaller models and faster training. In: International conference on machine learning, PMLR, Association for Computing Machinery(ACM) 2021, pp.10096–10106.
38.
GeZLiuSWangF, et al. Yolox: exceeding yolo series in 2021. arXiv preprint arXiv: 2107.08430, 2021.
