Search

Real-time what-if traffic prediction is crucial for decision making in intelligent traffic management and control. Although current deep learning methods demonstrate significant advantages in traffic prediction, they are powerless in what-if traffic prediction due to their nature of correlation-based. Here, we present a simple deep learning framework called TraffNet that learns the mechanisms of traffic generation for what-if prediction from vehicle trajectory data. First, we use a heterogeneous graph to represent the road network, allowing the model to incorporate causal features of traffic flows, such as Origin-Destination (OD) demands and routes. Next, we propose a method for learning segment representations, which involves modeling the process of assigning OD demands onto the road network. The learned segment representations effectively encapsulate the intricate causes of traffic generation, facilitating downstream what-if traffic prediction. Finally, we conduct experiments on synthetic datasets to evaluate the effectiveness of TraffNet. The code and datasets of TraffNet is available at https://github.com/mayunyi-1999/TraffNet_code.git.

The optimization of traffic signal control (TSC) is critical for an efficient transportation system. In recent years, reinforcement learning (RL) techniques have emerged as a popular approach for TSC and show promising results for highly adaptive control. However, existing RL-based methods suffer from notably poor real-world applicability and hardly have any successful deployments. The reasons for such failures are mostly due to the reliance on over-idealized traffic simulators for policy optimization, as well as using unrealistic fine-grained state observations and reward signals that are not directly obtainable from real-world sensors. In this paper, we propose a fully Data-Driven and simulator-free framework for realistic Traffic Signal Control (D2TSC). Specifically, we combine well-established traffic flow theory with machine learning to construct a reward inference model to infer the reward signals from coarse-grained traffic data. With the inferred rewards, we further propose a sample-efficient offline RL method to enable direct signal control policy learning from historical offline datasets of real-world intersections. To evaluate our approach, we collect historical traffic data from a real-world intersection, and develop a highly customized simulation environment that strictly follows real data characteristics. We demonstrate through extensive experiments that our approach achieves superior performance over conventional and offline RL baselines, and also enjoys much better real-world applicability., Comment: 15 pages, 6 figures

We present a novel framework, Action Progression Network (APN), for temporal action detection (TAD) in videos. The framework locates actions in videos by detecting the action evolution process. To encode the action evolution, we quantify a complete action process into 101 ordered stages (0\%, 1\%, ..., 100\%), referred to as action progressions. We then train a neural network to recognize the action progressions. The framework detects action boundaries by detecting complete action processes in the videos, e.g., a video segment with detected action progressions closely follow the sequence 0\%, 1\%, ..., 100\%. The framework offers three major advantages: (1) Our neural networks are trained end-to-end, contrasting conventional methods that optimize modules separately; (2) The APN is trained using action frames exclusively, enabling models to be trained on action classification datasets and robust to videos with temporal background styles differing from those in training; (3) Our framework effectively avoids detecting incomplete actions and excels in detecting long-lasting actions due to the fine-grained and explicit encoding of the temporal structure of actions. Leveraging these advantages, the APN achieves competitive performance and significantly surpasses its counterparts in detecting long-lasting actions. With an IoU threshold of 0.5, the APN achieves a mean Average Precision (mAP) of 58.3\% on the THUMOS14 dataset and 98.9\% mAP on the DFMAD70 dataset., Comment: Under Review. Code available at https://github.com/makecent/APN

Accurate hand joints detection from images is a fundamental topic which is essential for many applications in computer vision and human computer interaction. This paper presents a two stage network for hand joints detection from single unmarked image by using serial-parallel multi-scale feature fusion. In stage I, the hand regions are located by a pre-trained network, and the features of each detected hand region are extracted by a shallow spatial hand features representation module. The extracted hand features are then fed into stage II, which consists of serially connected feature extraction modules with similar structures, called "multi-scale feature fusion" (MSFF). A MSFF contains parallel multi-scale feature extraction branches, which generate initial hand joint heatmaps. The initial heatmaps are then mutually reinforced by the anatomic relationship between hand joints. The experimental results on five hand joints datasets show that the proposed network overperforms the state-of-the-art methods.

Introduction Non-recurrent congestion caused by traffic incident is difficult to predict but should be dealt with in a timely and effective manner to reduce its influence on road capacity reduction and enormous travel time loss. Influence factor analysis and reasonable prediction of traffic incident duration are important in traffic incident management to predict incident impacts and aid in the implementation of appropriate traffic operation strategies. The objective of this study is to conduct a thorough review and discusses the research evolution, mainly including the different phases of incident duration, data resources, and the various methods that are applied in the traffic incident duration influence factor analysis and duration time prediction. Methods In order to achieve the goal of this study, we presented a systematic review of traffic incident duration time estimation and prediction methods developed based on various data resource, methodologies etc. Results based on the previous studies, we analyse (i) Data resources and characteristics: different traffic incident time phases, data set size, incident types, duration time distribution, available data resources, significant influence factors and unobserved heterogeneity and randomness, (ii) traffic incident duration analysis methods, mainly including hazard-based duration model and regression and statistical tests, (iii) traffic incident duration prediction methods and evaluation of prediction accuracy. Conclusions After a comprehensive review of literature, this study identifies and analyses future challenges and what can be achieved in the future to estimate and predict the traffic incident duration time. Keywords: Incident duration analysis; Traffic incident duration prediction; Hazard-based duration model; Data mining; Influence factors

Real-time traffic flow prediction can not only provide travelers with reliable traffic information so that it can save people's time, but also assist the traffic management agency to manage traffic system. It can greatly improve the efficiency of the transportation system. Traditional traffic flow prediction approaches usually need a large amount of data but still give poor performances. With the development of deep learning, researchers begin to pay attention to artificial neural networks (ANNs) such as RNN and LSTM. However, these ANNs are very time-consuming. In our research, we improve the Deep Residual Network and build a dynamic model which previous researchers hardly use. We firstly integrate the input and output of the $i^{th}$ layer to the input of the $i+1^{th}$ layer and prove that each layer will fit a simpler function so that the error rate will be much smaller. Then, we use the concept of online learning in our model to update pre-trained model during prediction. Our result shows that our model has higher accuracy than some state-of-the-art models. In addition, our dynamic model can perform better in practical applications., Comment: 15 pages, 11 figures, 3 tables

Non-recurrent congestion caused by traffic incident is difficult to predict but should be dealt with in a timely and effective manner to reduce its influence on road capacity reduction and enormous travel time loss. Influence factor analysis and reasonable prediction of traffic incident duration are important in traffic incident management to predict incident impacts and aid in the implementation of appropriate traffic operation strategies. The objective of this study is to conduct a thorough review and discusses the research evolution, mainly including the different phases of incident duration, data resources, and the various methods that are applied in the traffic incident duration influence factor analysis and duration time prediction.In order to achieve the goal of this study, we presented a systematic review of traffic incident duration time estimation and prediction methods developed based on various data resource, methodologies etc.based on the previous studies, we analyse (i) Data resources and characteristics: different traffic incident time phases, data set size, incident types, duration time distribution, available data resources, significant influence factors and unobserved heterogeneity and randomness, (ii) traffic incident duration analysis methods, mainly including hazard-based duration model and regression and statistical tests, (iii) traffic incident duration prediction methods and evaluation of prediction accuracy.After a comprehensive review of literature, this study identifies and analyses future challenges and what can be achieved in the future to estimate and predict the traffic incident duration time.

Xia Yang,1 Yali Xu,1 Jianmin Jin,1 Ruimin Li,1 Xiaofang Liu,1 Yongchang Sun1,2 1Department of Respiratory Medicine, Beijing Tongren Hospital, Capital Medical University, 2Department of Respiratory and Critical Care Medicine, Peking University Third Hospital, Beijing, People’s Republic of China Background and purpose: Bronchiectasis revealed by high-resolution computed tomography (HRCT) is common in chronic obstructive pulmonary disease (COPD), but the causes and risk factors remain to be determined. Chronic rhinosinusitis (CRS) is closely associated with bronchiectasis or COPD, but whether it is associated with comorbid bronchiectasis in COPD (COPD-Bx) is unknown.Patients and methods: Patients with stable COPD were enrolled consecutively and ­evaluated for the presence of CRS by questionnaire and paranasal sinus computed tomography. The presence and severity of bronchiectasis on lung HRCT were evaluated by the Smith and severity scores. COPD symptoms were evaluated by COPD Assessment Test (CAT) and Modified British Medical Research Council Questionnaire. The sputum cell differentials and concentrations of interleukin (IL)-6, IL-8, IL-5, matrix metalloproteinases-9 (MMP-9), and tissue inhibitor of matrix metalloproteinases-1 were measured.Results: We enrolled 136 patients with stable COPD, of which 66 (48.5%) were diagnosed with CRS according to the European Position Paper on Rhinosinusitis and Nasal Polyps (EP3OS) criteria. The prevalence of bronchiectasis was 57.6% in patients with CRS, but 37.1% in those without CRS (P=0.017). COPD-Bx patients with CRS showed a significantly higher severity score of bronchiectasis than those without CRS (P=0.034). COPD patients with CRS had a higher percentage of eosinophils, higher levels of IL-8, IL-6, and MMP-9 in sputum as compared to those without CRS. In COPD-Bx patients with CRS, the percentage of eosinophils and the levels of IL-6 and MMP-9 in sputum were increased as compared to those without CRS. In all

Although Set-top-Boxes(STBs) are widely deployed today to connect a media source to a display (traditionally a television set), the market is changing due to the introduction of Internet Protocol Television, Over-the-Top streaming devices, gaming console, home theater Personal Computer, smart TV, etc. There is an evolving concept of a Home Media Center (HMC). This HMC provides consumers with an integrated home media environment and experience. This thesis explores the transition from STBs to HMCs. The specific questions that this thesis project answers are: What will a future HMC look like? What will its functions be? What interfaces and protocols will it use? Who will make these HMCs? How can STB vendors evolve to be HMC vendors or will they simply cease to exist? This thesis project designed and evaluated a hypothetical HMC prototype based upon current technology trends and user expectations. This prototype was used with 68 volunteers to identify and prioritize the most important features that a HMC should provide. Based upon the most important of these features a conceptual HMC prototype is designed to define a HMC product roadmap for 1, 3, and 5 years. This roadmap is used to project the economic impact of HMCs on the current STB industry. This economic analysis considers Sweden as the target market. This thesis could be used by current STB vendors to define their own company specific roadmaps to support their transition to the future HMC market., Aven digitalboxar är spridda i dag för att ansluta en mediakälla till en bildskärm (traditionellt en TV) är marknaden förändras på grund av införandet av Internet Protocol Television, Over-the-Top streaming anordningar, spelkonsol, hem teater Personal Computer, smarta TV osv. Det finns en framväxande begreppet Home Media Center (HMC). Detta HMC ger konsumenterna ett integrerat hem mediemiljö och erfarenhet. Denna avhandling utforskar övergången från digitalboxar till HMC. De specifika frågor som detta examensarbete skall besvara är: Vad kommer en framtida HMC se ut? Hur kommer dess funktioner att bli? Vilka gränssnitt och protokoll kommer den att använda? Vem kommer att göra dessa HMC? Hur kan digitalboxar leverantörer utvecklas vara HMC säljare eller kommer de helt enkelt att upphöra att existera? Detta examensarbete designar och utvärderar en hypotetisk HMC prototyp baserad på nuvarande tekniktrender som användarnas förväntningar studie. Denna prototyp användes med 68 frivilliga att identifiera och prioritera de viktigaste funktionerna som en HMC bör ge. Baserat på de viktigaste av dessa funktioner en konceptuell HMC prototyp kommer att utformas för att definiera en HMC produkt färdplan för en, 3 och 5 år. Färdplanen kommer att användas för att projicera de ekonomiska konsekvenserna av HMC på den aktuella digitalboxar industrin. Denna ekonomiska analysen kommer att överväga Sverige som målgrupp. Denna avhandling kan användas av nuvarande digitalboxar leverantörer för att definiera sina egna företagsspecifika färdplaner för att stödja övergången till den framtida HMC marknaden.

Risk measures are the core indicator of risk management and a proper risk assessment model is essential for successful financial institutions. Value at Risk and Expected Shortfall are the two most popular and acceptable risk measurement methods presently employed to assess risks in the financial market. In the past few years, researchers have attempted to demonstrate that Expected Shortfall performs better against the traditional Value at Risk method. However, the lack of elicitability and difficult backtesting of this method suggest that the popularisation of ES might be gradual. This thesis will present a comparison of these two methods not only from a traditional perspective, such as the measurement of tail risk, but also form the perspective of risk capital requirement. Through Historical Simulation and Filtered Historical Simulation, it concludes that switching from Value at Risk to Expected Shortfall method would reduce risk capital requirement and enhance financial leverage of organisations. Additionally, this research also combines macroeconomic elements, the financial market and central banks, and analyses the influence of a positive leverage shock on the macro-economy through a Global Vector Autoregression model.

Risk measures are the core indicator of risk management and a proper risk assessment model is essential for successful financial institutions. Value at Risk and Expected Shortfall are the two most popular and acceptable risk measurement methods presently employed to assess risks in the financial market. In the past few years, researchers have attempted to demonstrate that Expected Shortfall performs better against the traditional Value at Risk method. However, the lack of elicitability and difficult backtesting of this method suggest that the popularisation of ES might be gradual. This thesis will present a comparison of these two methods not only from a traditional perspective, such as the measurement of tail risk, but also form the perspective of risk capital requirement. Through Historical Simulation and Filtered Historical Simulation, it concludes that switching from Value at Risk to Expected Shortfall method would reduce risk capital requirement and enhance financial leverage of organisations. Additionally, this research also combines macroeconomic elements, the financial market and central banks, and analyses the influence of a positive leverage shock on the macro-economy through a Global Vector Autoregression model.