Due to the unavailability of Vehicle-to-Infrastructure (V2I) communication in current transportation systems, Traffic Light Detection (TLD) still is considered an important module in autonomous vehicles and Driver Assistance Systems (DAS). Considering low flexibility and accuracy of vision-based heuristic algorithms and high power consumption of deep learning-based methods, we propose a lightweight and real-time traffic light detector for the autonomous vehicle platform on the basis of previous studies. Our model consists of a heuristic candidate region selection module to identify all possible traffic lights, and a lightweight Convolution Neural Network (CNN) classifier to classify the results obtained. Offline simulations on the GPU server with the collected dataset and several public datasets show that our model achieves higher average accuracy and less time consumption. By integrating our detector module on NVidia Jetson TX1/TX2, we conduct on-road tests on two full scale self-driving vehicle platforms (a car and a bus) in normal traffic conditions. The final model can achieve average detection accuracy of 99.3% (mRttld) and 99.7% (Rttld) at 10Hz on TX1 and TX2, respectively. The on-road tests also show that our traffic light detection module can achieve < ±1.5m errors at stop lines when working with other self-driving modules.

In order to improve trajectory tracking accuracy, a reinforcement learning method was employed to address the trajectory tracking task in autonomous driving. There are many conveniences and advantages in theories, methods and tools utilizing reinforcement learning to solve trajectory tracking control problems. To teach an intelligent agent effective driving skills, the structured road information in the urban environment was extracted to generate an accurate reference trajectory. Then real-world scenarios were modeled to build a simulation environment for training. To effectively train the intelligent driving agent, Imitation Learning was firstly employed to teach the agent primary driving skills. Afterwards, Reinforcement Learning was adopted to optimize the agent’s driving policy. After the intelligent driving agent was well trained in the simulator, the tracking experiments were conducted in the simulator and the real-world scenarios. The proposed method was compared with base-line methods of geometric tracking, optimization-based tracking and learning-based tracking. The experimental results demonstrated that Reinforcement-Tracking can achieve accurate trajectory tracking performance and even exceed the accuracy of most baseline methods.

EEG-based emotion recognition is a task that uses scalp-EEG data to classify the emotion states of humans. The study of EEG-based emotion recognition can contribute to a large spectrum of application fields including healthcare and human–computer interaction. Recent studies in neuroscience reveal that the brain regions and their interactions play an essential role in the processing of different stimuli and the generation of corresponding emotional states. Nevertheless, such regional interactions, which have been proven to be critical in recognizing emotions in neuroscience, are largely overlooked in existing machine learning or deep learning models, which focus on individual channels in brain signals. Motivated by this, in this paper, we present RGNet, a model that is designed to learn the regional level representation of EEG signal for accurate emotion recognition. Specifically, after applying preprocessing and feature extraction techniques on raw signals, RGNet adopts a novel region-wise encoder to extract the features of channels located within each region as input to compute the regional level features, enabling the model to effectively explore the regional functionality. A graph is then constructed by considering each region as a node and connections between regions as edges, upon which a graph convolutional network is designed with spectral filtering and learned adjacency matrix. Instead of focusing on only the spatial proximity, it allows the model to capture more complex functional relationships. We conducted experiments from the perspective of region division strategies, region encoders and input feature types. Our model has achieved $98.64\%$ and $99.33\%$ for Deap and Dreamer datasets, respectively. The comparison studies show that RGNet outperforms the majority of the existing models for emotion recognition from EEG signals.

Introduction Salt intake is very high in China, with ≈80% being added by the consumers. It is difficult to reduce salt in such settings. Our previous study (School-based Education programme to reduce Salt(School-EduSalt)) demonstrated that educating schoolchildren, who then instructed their families to reduce the amount of salt used at home, is effective in lowering salt intake in both children and adults. Our team also developed an app called ‘KnowSalt’, which could help individuals to estimate their salt intake and the major sources of salt in the diet. Building on School-EduSalt and KnowSalt, we propose to develop a new app (AppSalt) focusing on salt reduction through education, target setting, monitoring, evaluation, decision support and management to achieve a progressive lower salt intake for long term. To evaluate the effectiveness of the AppSalt programme, we will carry out a cluster randomised controlled trial. Methods and analysis We will recruit 54 primary schools from urban and rural areas of three provinces in China. A total of 594 children aged 8–9 years and 1188 adult family members will be randomly selected for evaluation. After baseline assessment, schools will be randomly allocated to either the intervention or control group. Children in the intervention group will be taught, with support of AppSalt, about salt reduction and assigned homework to get the whole family involved in the activities to reduce salt consumption. The duration of the intervention is two school terms (ie, 1 year). The primary outcome is the difference between the intervention and control group in the change of salt intake as measured by 24-hour urinary sodium. Ethics and dissemination The study has been approved by Queen Mary Research Ethics Committee and Peking University Health Science Centre IRB. Results will be disseminated through presentations, publications and social media. Trial registration number ChiCTR1800017553.