Multiple image hiding aims to hide multiple secret images into a single cover image, and then recover all secret images perfectly. Such high-capacity hiding may easily lead to contour shadows or color distortion, which makes multiple image hiding a very challenging task. In this paper, we propose a novel multiple image hiding framework based on invertible neural network, namely DeepMIH. Specifically, we develop an invertible hiding neural network (IHNN) to innovatively model the image concealing and revealing as its forward and backward processes, making them fully coupled and reversible. The IHNN is highly flexible, which can be cascaded as many times as required to achieve the hiding of multiple images. To enhance the invisibility, we design an importance map (IM) module to guide the current image hiding based on the previous image hiding results. In addition, we find that the image hidden in the high-frequency sub-bands tends to achieve better hiding performance, and thus propose a low-frequency wavelet loss to constrain that no secret information is hidden in the low-frequency sub-bands. Experimental results show that our DeepMIH significantly outperforms other state-of-the-art methods, in terms of hiding invisibility, security and recovery accuracy on a variety of datasets.

Early fault detection technique is crucial to reduce the machine downtime and has high impact on a wide variety of industrial applications. However, early fault detection is still subject to the following challenges: 1) extracting features from incipient fault signals; 2) detecting anomalies with considering sequential data correlation; and 3) enhancing the reliability of fault alarm. In this paper, we introduce a novel deep-structured framework to solve the early fault detection problem. First, the system variation is measured with the deviation value generated by a current feature extraction model using deep neural network (DNN) and a distribution estimator based on the long short-term memory (LSTM) network. DNN has the ability of representing a complicated and intrinsic distribution for data, which is suitable for handling the early fault data masked by heavy noise, and LSTM is able to discover temporal dependencies in high-dimensional sequential data, which allows distribution estimator making use of previous context information as well as makes the distribution estimator more robust to warp along the time axis. Second, a circular indirect alarm assessment strategy is designed for collecting deviation values and confirming the fault appearance only when a specified confidence level is reached. Experimental results on the typical real-world bearing data sets demonstrate the effectiveness and the reliability of our model.

Recent years have witnessed remarkable achievements in video-based action recognition. Apart from traditional frame-based cameras, event cameras are bio-inspired vision sensors that only record pixel-wise brightness changes rather than the brightness value. However, little effort has been made in event-based action recognition, and large-scale public datasets are also nearly unavailable. In this paper, we propose an event-based action recognition framework called EV-ACT . The Learnable Multi-Fused Representation (LMFR) is first proposed to integrate multiple event information in a learnable manner. The LMFR with dual temporal granularity is fed into the event-based slow-fast network for the fusion of appearance and motion features. A spatial-temporal attention mechanism is introduced to further enhance the learning capability of action recognition. To prompt research in this direction, we have collected the largest event-based action recognition benchmark named THU^E-ACT-50 and the accompanying THU^E-ACT-50-CHL dataset under challenging environments, including a total of over 12,830 recordings from 50 action categories, which is over 4 times the size of the previous largest dataset. Experimental results show that our proposed framework could achieve improvements of over 14.5%, 7.6%, 11.2%, and 7.4% compared to previous works on four benchmarks. We have also deployed our proposed EV-ACT framework on a mobile platform to validate its practicality and efficiency.

Unmanned Aerial Vehicle (UAV) technology has been widely applied in both military and civilian applications. Recent researches on UAV systems feature in the dramatic augment of the variety and number of equipped sensors, which results in such an issue that multiple UAVs cannot afford to handle the big data generated by a range of sensors in the air. Considering this practical problem, in this paper, we propose a cloud-based UAV system which incorporates the computing capability of the terrestrial cloud into the UAV systems. Relying on proposed cloud-based UAV system, one critical theoretic issue is how to acquire the big data generated by the sensors while guaranteeing a stable operation state of the system. First, we analyze the cloud-based system's on-demand service ability as well as its impact on UAVs' control procedure. Second, the UAV cloud control system is modeled as a network control system. Moreover, the stable condition of the UAV cloud control system is derived, which reveals the relationship between the acquisition rate of sensor data and the stability of the cloud-based UAV system. Finally, simulations are conducted to verify the effectiveness of our theoretical analysis.

Traditional frame-based video frame interpolation (VFI) methods rely on the linear motion assumption and brightness invariance assumption, which may lead to fatal errors confronting the scenarios with high-speed motions. To tackle the above challenge, inspired by the advantages of event cameras on asynchronously recording brightness changes at each pixel, we propose a Fast-Slow joint synthesis framework for event-enhanced high-speed video frame interpolation, named SuperFast , in this paper, which can generate high frame rate (5000 FPS, 200× faster) video from the input low frame rate (25 FPS) video and the corresponding event stream. In our framework, the task is divided into two sub-tasks, i.e., video frame interpolation for the contents with and without high-speed motions, which are tackled by two corresponding branches, i.e., the fast synthesis pathway and the slow synthesis pathway. The fast synthesis pathway leverages a spiking neural network to encode the input event stream, and combines boundary frames to generate intermediate results through synthesis and refinement, targeting on contents with high-speed motions. The slow synthesis pathway stacks the two input boundary frames and the event stream to synthesize intermediate results, focusing on relatively slow-motion contents. Finally, a fusion module with a comparison loss is utilized to generate the final video frame interpolation results. We also build a hybrid visual acquisition system containing an event camera and a high frame rate camera, and collect the first 5000 FPS H igh- S peed E vent-enhanced V ideo frame I nterpolation (THU $^{\text{HSEVI}}$ ) dataset. To evaluate the performance of our proposed framework, we have conducted experiments on our THU $^{\text{HSEVI}}$ dataset and the existing HS-ERGB dataset. Experimental results demonstrate that our proposed framework can achieve state-of-the-art 200× video frame interpolation performance under high-speed motion scenarios.