Search

Unsupervised video object segmentation (UVOS) aims at segmenting the primary objects in videos without any human intervention. Due to the lack of prior knowledge about the primary objects, identifying them from videos is the major challenge of UVOS. Previous methods often regard the moving objects as primary ones and rely on optical flow to capture the motion cues in videos, but the flow information alone is insufficient to distinguish the primary objects from the background objects that move together. This is because, when the noisy motion features are combined with the appearance features, the localization of the primary objects is misguided. To address this problem, we propose a novel reciprocal transformation network to discover primary objects by correlating three key factors: the intra-frame contrast, the motion cues, and temporal coherence of recurring objects. Each corresponds to a representative type of primary object, and our reciprocal mechanism enables an organic coordination of them to effectively remove ambiguous distractions from videos. Additionally, to exclude the information of the moving background objects from motion features, our transformation module enables to reciprocally transform the appearance features to enhance the motion features, so as to focus on the moving objects with salient appearance while removing the co-moving outliers. Experiments on the public benchmarks demonstrate that our model significantly outperforms the state-of-the-art methods. Code is available at https://github.com/OliverRensu/RTNet.

This paper tackles the task of Few-Shot Video Object Segmentation (FSVOS), i.e., segmenting objects in the query videos with certain class specified in a few labeled support images. The key is to model the relationship between the query videos and the support images for propagating the object information. This is a many-to-many problem and often relies on full-rank attention, which is computationally intensive. In this paper, we propose a novel Domain Agent Network (DAN), breaking down the full-rank attention into two smaller ones. We consider one single frame of the query video as the domain agent, bridging between the support images and the query video. Our DAN allows a linear space and time complexity as opposed to the original quadratic form with no loss of performance. In addition, we introduce a learning strategy by combining meta-learning with online learning to further improve the segmentation accuracy. We build a FSVOS benchmark on the Youtube-VIS dataset and conduct experiments to demonstrate that our method outperforms baselines on both computational cost and accuracy, achieving the state-of-the-art performance. Code is available at https://github.com/scutpaul/DANet.

This paper presents a constant current regulation for discontinuous conduction mode primary-side controlled flyback converter. It consists of a demagnetization time detector and a peak current threshold generator. Based on peak current control, the regulation is achieved by fixing the product of the peak current and the ratio of the demagnetization time to the switching period constant. Using a reference voltage and the demagnetization time provided by the demagnetization time detector, the peak current threshold generator generates the threshold signal, which is relative to the ratio of the demagnetization time to the switching period. The circuit is simulated in software of Spectre based on the nuvoton 0.35um BCD process and a control IC has been fabricated. Experimental results of a 7V/1.5A circuit prototype shows that the constant output current precision is within 5.4% when the input voltage changes from 100V to 320V and while the load changes from 1.6H to 4.5H, the output current changes by 4.1%.

This paper proposes four newly-designed antenna configurations for the applications of short-distance wireless power transfer. The antennas are developed from a traditional circular loop antenna, which have a structural feature in common that all of them are in enclosed shape. Four proposed designs aim to improve the transfer efficiency. In this work, the transmitting antenna faces to the receiving antenna with a distance of 12 cm in the wireless power transfer system. Results show that two of them have better performance of transferring power than the other two and the conventional design.

With the further market reform of electric power industry, customer-oriented marketing concept has been more and more important in electric power enterprises. After doing research on customer service data of an electric power enterprise, we pick up some key factors which might be related to customer satisfaction, and utilize a developed association rule model to mine association rules between these key factors and customer satisfaction, which reflects the close relationship between them. Furthermore, we design metrics that can reflect the changing trend of rule confidence, which helps business officers have an insight into how these factors influence customer satisfaction as time goes on, to effectively improve customer service quality and then win praise of customers. As practical experiments prove, this method can fulfill the goal very well.