Search

Estimation of absolute temperature distributions is crucial for many thermal processes in the nonlinear distributed parameter systems, such as predicting the curing temperature distribution of the chip, the temperature distribution of the catalytic rod, and so on. In this work, a spatiotemporal model based on the Karhunen-Loève (KL) decomposition, the multilayer perceptron (MLP), and the long short-term memory (LSTM) network, named KL-MLP-LSTM, is developed for estimating temperature distributions with a three-step procedure. Firstly, the infinite-dimensional model is transformed into a finite-dimensional model, where the KL decomposition method is used for dimension reduction and spatial basis functions extraction. Secondly, a novel MLP-LSTM hybrid time series model is constructed to deal with the two inherently coupled nonlinearities. Finally, the spatiotemporal temperature distribution model can be reconstructed through spatiotemporal synthesis. The effectiveness of the proposed model is validated by the data from a snap curing oven thermal process. Satisfactory agreement between the results of the current model and the other well-established model shows that the KL-MLP-LSTM model is reliable for estimating the temperature distributions during the thermal process.

The main challenge in video salient object detection is how to model object motion and dramatic changes in appearance contrast. In this work, we propose an attention embedded spatio-temporal network (ASTN) to adaptively exploit diverse factors that influence dynamic saliency prediction within a unified framework. To compensate for object movement, we introduce a flow-guided spatial learning (FGSL) module to directly capture effective motion information in the form of attention based on optical flows. However, optical flow represents the motion information of all moving objects, including movements of non-salient objects caused by large camera motion and subtle changes in background. Therefore, using the flow-guided attention map alone causes the spatial saliency to be influenced by all moving objects rather than just the salient objects, resulting in unstable and temporally inconsistent saliency maps. To further enhance the temporal coherence, we develop an attentive bidirectional gated recurrent unit (AB-GRU) module to adaptively exploit sequential feature evolution. With this AB-GRU, we can further refine the spatiotemporal feature representation by incorporating an accommodative attention mechanism. Experimental results demonstrate that our model achieves superior empirical performance on video salient object detection. Moreover, an experiment on the extended application to unsupervised video object segmentation further demonstrates the generalization ability and stability of our proposed method.

In this paper, a novel neural network named CRSum for the video summarization task is proposed. The proposed network integrates feature extraction, temporal modeling, and summary generation into an end-to-end architecture. Compared with previous work on this task, the proposed method owns three distinctive characteristics: 1) it for the first time leverages convolutional recurrent neural network for simultaneously modeling spatial and temporal structure of video for summarization; 2) thorough and delicate features of video are obtained in the proposed architecture by trainable three-dimension convolutional neural networks and feature fusion; and 3) a new loss function named Sobolev loss is defined, aiming to constrain the derivative of sequential data and exploit potential temporal structure of video. A series of experiments are conducted to prove the effectiveness of the proposed method. We further analyze our method from different aspects by well-designed experiments.

With the rapid progress of urbanization, predicting citywide crowd flows has become increasingly significant in many fields, such as traffic management and public security. However, influenced by the complex spatiotemporal relations in raw data and other factors, such as events and weather, obtaining a precise prediction is challenging. Some previous works attempted to address this problem using various ways, such as autoregressive integrated moving average, vector auto-regression and some deep learning models. However, seldom can these methods comprehensively capture the spatiotemporal correlations. In this paper, we propose a novel spatio-temporal prediction model that is based on densely connected convolutional networks and attention long short-term memory (ST-DCCNAL), to simultaneously predict the inflow and outflow of the crowds in regions divided within a specific city. The ST-DCCNAL model consists of three parts: spatial part, external factors part and temporal part. In the spatial part, we employ densely connected convolutional networks to extract spatial characteristics at different levels. The external factors part utilizes a fully connected network to extract features from auxiliary information. In the last part, an attention-based long short-term memory module is leveraged to capture the temporal pattern. To demonstrate the practicality and effectiveness of the proposed model, we evaluate it using two separate real-world datasets of taxis in Beijing and bikes in New York. The experimental results confirm that the performance of our model is better than that of other baseline methods.