Motion magnification techniques extend the perception scope of naked eyes to tiny variations, such as the muscle tremors and the mechanism vibrations. Current approaches are developed respectively from Lagrangian and Eulerian perspectives. However, these approaches either require complex computation or cannot distinguish subtle variations from noise. This paper proposes a novel motion magnification approach, fusing Lagrangian and Eulerian methods via multi-task learning. The approach is mainly developed from Eulerian methods for efficient inference and introduces optical flow from Lagrangian methods for precise motion perception. To optimize the training process, homoscedastic uncertainty is introduced to balance these tasks. To overcome the lack of real magnified images, this paper establishes a synthetic dataset by real images selected from public datasets. The dataset simulates tiny and magnified motions by image preprocessing and affine transformations. Through qualitative and quantitative experiments, the proposed approach outperforms previous ones with few artifacts and strong robustness to magnification factors, motion magnitude, and noise disturbance. Additionally, the optical flow subnet is evaluated by public benchmarks to demonstrate its motion extraction capacity and assistance offered to motion magnification.

Vision-based vibration measurement technology has received extensive attention due to its advantages of non-contact, high spatial resolution, and no-load effect. However, with the complexity of measurement objects and measurement tasks, the existing visual measurement technology is gradually showing greater limitations. Specifically, due to the uncertainty of actual working conditions, not all pixels in the field of view can measure vibration. Therefore, the selection of measurement points needs to rely on prior structural information and artificial experience. Frequent manual point selection tests bring a lot of resource consumption, which greatly reduces the automation degree of visual vibration measurement. This paper focuses on an intelligent area localization method for vibration measurement of rotating machine vision and designs a deep learning-based vibration measurement area localization framework to directly feedback all reliable measurement pixels from image data, which is called the VMAL framework. Firstly, the sub-pixel physical feature information associated with vibration in the data is analyzed through an unsupervised image decomposition network, and then a regularized regional localization network is used to cluster and output reliable regional pixels. Experimental results on a medium-sized single-span rotor platform verify the effectiveness of the proposed method.

Recently, deep transfer learning (TL) has successfully addressed the problem of fault diagnosis under variable operating conditions. Existing methods default that the source and target domains have the same label space, and solve distribution discrepancy problem under different working conditions by aligning their feature distributions. However, in the practical industry, is unlikely to guarantee the health conditions of the target domain data are consistent with the source domain. Therefore, industrial applications usually face the challenge of more difficult partial domain diagnosis scenarios. In this paper, a deep partial domain adaptation network based on a balanced alignment constraint strategy is proposed to realize cross-domain diagnosis. The proposed method combines balanced augmentation and subdomain alignment, which can effectively facilitate the positive transfer of shared categories. Meanwhile, the conditional entropy minimization is introduced to encourage the predictions of target domain samples with high confidence. The experimental results on the rolling bearing dataset verify the effectiveness and feasibility of the proposed method in handling the actual partial domain fault diagnosis problem.

This paper introduces a novel method for real-time 2D hand pose estimation from monocular color images, which is named as SRHandNet. Existing methods can not time efficiently obtain appropriate results for small hand. Our key idea is to simultaneously regress the hand region of interests (RoIs) and hand keypoints for a given color image, and iteratively take the hand RoIs as feedback information for boosting the performance of hand keypoints estimation with a single encoder-decoder network architecture. Different from previous region proposal network (RPN), a new lightweight bounding box representation, which is called region map, is proposed. The proposed bounding box representation map together with hand keypoints heatmaps are combined into the unified multi-channel feature maps, which can be easily acquired with only one forward network inference and thus improve the runtime efficiency of the network. Our proposed SRHandNet can run at 40fps for hand bounding box detection and up to 30fps accurate hand keypoints estimation under the desktop environment without implementation optimization. Experiments demonstrate the effectiveness of the proposed method. State-of-the-art results are also achieved out competing all recent methods.

We present a cascaded convolutional neural network for 2D hand pose estimation from single in-the-wild RGB images. Inspired by the commonly used silhouette information in the generative pose estimation approaches, we build the cascaded network with two stages, including mask prediction stage as well as pose estimation stage. We find that the two stages network architecture for end-to-end training could benefit from each other for detecting the hand mask and 2D pose. To further improve the hand pose detection accuracy, we contribute a new RGB hand dataset named OneHand10K, which contains 10K RGB images. Each image contains one single hand. We manually obtain the segmented mask and labeled keypoints for guided learning. We hope that this dataset will be a benchmark and encourage more people to conduct research on this challenging topic. Experiments on the validation dataset have demonstrated the superior performance of the proposed cascaded convolutional neural network.

Back-EMF zero crossing-based sensorless motor driving method suffers from a starting problem since there is no back-EMF at standstill, while the V/f or I/f based method is parameter sensitive. Without position sensors, smoothly driving a nonsalient BLDC motor from standstill is difficult. To solve this problem, the relationship between preposition current, energizing time, and transient speed in startup process is analyzed. Based on this relationship, starting from standstill and directly switching to self-synchronization stage can be realized without external-synchronization process. In addition, to lower the back-EMF zero-crossing points (ZCPs) detection speed, the line voltages difference ZCPs detection circuit is designed, and a novel phase error correction method based on a neurofuzzy controller is presented to provide accurate commutation point in the high-speed range. The experimental results show that reliable startup and accurate commutation can be guaranteed by the proposed strategy.