Traffic light detection is a key module in the autonomous driving system to enhance the interactions between drivers and unmanned vehicles. In recent studies, deep neural networks are widely used for traffic light detection and resource/power consumption is a major concern for model deployment in vehicular edge devices. This paper proposes a novel light-weight deep CNN model that integrates the multi-backbone of state-of-the-art architectures for the self-driving traffic light detection. The MBBNet (Multi-BackBone Network) consists of three common convolutional backbones, i.e., the normal, residual and highway (DenseNet) convolutional modules. Simple ensemble of those backbones may incur high computational load. Therefore, channel compression is adopted to control the model parameters, while guaranteeing the accuracy for mobile and embedded hardware. Evaluation of a dataset collected from real road conditions demonstrate the robustness of our detection system, and it achieves higher accuracy (accuracy > 0.94 and $Average\_IOU>74.05\%$) for self-driving buses. In terms of resource consumption, the trained model size is 1.35 MB, and can process high-resolution images (1280 × 960) at 14 FPS (frames per second) on low-power edge devices.

交通信号灯检测是自动驾驶系统中的关键模块，可增强驾驶员与无人驾驶车辆之间的互动。在最近的研究中，深度神经网络已广泛用于交通信号灯检测，并且资源/功耗是车辆边缘设备中模型部署的主要关注点。本文提出了一种新颖的轻量级深度CNN模型，该模型集成了用于自动驾驶交通灯检测的最新架构的多主干。MBBNet（多骨干网络）由三个常见的卷积主干组成，即普通，残差和高速公路（DenseNet）卷积模块。这些骨干的简单集成可能会导致高计算量。因此，采用通道压缩来控制模型参数，同时保证移动和嵌入式硬件的准确性。对从实际道路状况收集的数据集的评估证明了我们检测系统的鲁棒性，并且可以实现更高的准确性（精度 > 0.94和$\mathrm{一种}vË\mathrm{[R}\mathrm{一种}GË\_\mathrm{一世}Øü>74.05％$）的自动驾驶巴士。在资源消耗方面，经过训练的模型大小为1.35 MB，并且可以在低功率边缘设备上以14 FPS（每秒帧）的速度处理高分辨率图像（1280×960）。