Cooperative perception (CP) has emerged as an important research topic in connected and automated transportation systems, significantly enhancing situational awareness in challenging conditions such as limited communication bandwidth, harsh weather, and poor lighting. By sharing and aggregating data from multiple sources—including vehicles, roadside units (RSUs), and other infrastructure—CP addresses critical issues like occlusion and long-range perception in complex environments. This article provides a comprehensive review of existing CP methodologies, evaluating their effectiveness across diverse scenarios and fusion layers. We investigate approaches ranging from early to late collaboration strategies, highlighting how multimodal sensing at the network level can improve detection accuracy, system robustness, and adaptability. Additionally, we review publicly available datasets, identifying key gaps in their coverage of heterogeneous sensors, adverse conditions, and large-scale spatial scenarios. Finally, we propose actionable future research directions, including advanced algorithms for adverse environments, adaptive fusion strategies for heterogeneous sensors, and the integration of emerging technologies such as high-definition (HD) maps, large language models (LLMs), and end-to-end frameworks. These advancements aim to ensure reliability, interoperability, and scalability for real-world connected and automated vehicle (CAV) applications.
In recent years, the safety of vulnerable road users (VRUs), such as pedestrians, cyclists, and micro-mobility users, has become an increasingly significant concern in urban transportation systems worldwide. Reliable and accurate detection of VRUs is essential for effective safety protection. This survey explores the techniques and methodologies used to detect VRUs, ranging from conventional methods to state-of-the-art (SOTA) approaches, with a primary focus on infrastructure-based detection. This study synthesizes findings from recent research papers and technical reports, emphasizing sensor modalities such as cameras, LiDAR, and RADAR. Furthermore, the survey examines benchmark datasets used to train and evaluate VRU detection models. Alongside innovative detection models and sufficient datasets, key challenges and emerging trends in algorithm development and dataset collection are also discussed. This comprehensive overview aims to provide insights into current advancements and inform the development of robust and reliable roadside detection systems to enhance the safety and efficiency of VRUs in modern transportation systems.
Whilst Deep Neural Networks (DNNs) have been developing swiftly, most of the research has been focused on videos based on RGB frames. RGB data has been traditionally optimised for human vision and is a highly re-elaborated and interpolated version of the collected raw data. In fact, the sensor collects the light intensity value per pixel, but an RGB frame contains 3 values, for red, green, and blue colour channels. This conversion to RGB requires computational resource, time, power, and increases by a factor of three the amount of output data. This work investigates DNN based detection using (for training and evaluation) Bayer frames, generated from a benchmarking automotive dataset ( i.e . KITTI dataset). A Deep Neural Network (DNN) is deployed in an unmodified form, and also modified to accept only single channel frames, such as Bayer frames. Eleven different re-trained versions of the DNN are produced, and cross-evaluated across different data formats. The results demonstrate that the selected DNN has the same accuracy when evaluating RGB or Bayer data, without significant degradation in the perception (the variation of the Average Precision is <1%). Moreover, the colour filter array position and the colour correction matrix do not seem to contribute significantly to the DNN performance. This work demonstrates that Bayer data can be used for object detection in automotive without significant perception performance loss, allowing for more efficient sensing-perception systems.
11 August 2025 in IEEE Robotics and Automation Letters
Dilemma zones at signalized intersections present a commonly occurring yet unsolved challenge in traffic safety. The onsets of yellow-light prompts varied responses from drivers: some may brake abruptly, compromising ride comfort, while others may accelerate, increasing the likelihood of red-light violations and potential safety hazards. This heterogeneity in drivers' stop-or-go (SoG) decisions stems from the surrounding traffic conditions, vehicle states, and individual driver characteristics. Consequently, identifying personalized driving behaviors and integrating them into advanced driver assistance systems (ADAS) to mitigate the dilemma zone issues poses a compelling scientific question. However, there are no open-source datasets specifically designed for studying driving behaviors (especially personalized behaviors) in dilemma zones. To fill these gaps, this study leverages a game engine-based (i.e., CARLA-enabled) driving simulator to collect high-resolution vehicle trajectories, traffic signal phase and timing information, and SoG decisions from twenty drivers across various dilemma zone scenarios. The dataset is publicly available on GitHub , facilitating in-depth analysis and modeling of personalized driving behaviors in dilemma zones. Furthermore, we propose a personalized convolutional neural network (CNN)-based model by incorporating the personalized information, including distance to stop-line, average speed, and average acceleration at the moment of SoG decision-making, and the probability of choosing the “Go” decision, to predict SoG decisions and decision times. The proposed model improves the accuracy of the SoG decision prediction by 5.9% and reduces the mean squared error (MSE) of decision time predictions by 18.8% compared to the Baseline CNN model without differentiating individual driving behaviors.
1 November 2026 in IEEE Intelligent Transportation Systems Magazine
Whilst Deep Neural Networks have been developing swiftly, most of the research has been focused on RGB image. This type of image has been traditionally optimised for human vision. However, RGB data is a highly re-elaborated and interpolated version of the collected raw data (i.e. the sensor collects one value per pixel), but an RGB image for human viewing contains 3 values, for red, green, and blue. This processing through the ISP (Image Signal Processing) requires computational resource, time, power and obviously increases by a factor of three the amount of output data. This work investigates Deep Neural Network based detection using (for training and evaluation) Bayer data, generated in different ways, from a benchmarking automotive dataset (i.e. KITTI dataset). A Deep Neural Network (DNN) is deployed in unmodified form, and also modified to accept only single field images, such as Bayer frames. Eleven different re-trained version of the DNN are produced, and cross-evaluated across different data formats. The results demonstrate that the selected DNN has the same accuracy when evaluating RGB or Bayer data, without significant degradation in the perception (the variation of the Average Precision is <1%). Moreover, the colour filter array position and the colour correction matrix do not seem to contribute significantly to the DNN performance. This work demonstrates that Bayer data can be used for object detection in automotive without significant performance loss, and that the processing currently used in ISP can be avoided, allowing for more efficient sensing-perception systems.