Search

Diabetes mellitus is prevalent among women of reproductive age, and many women are left undiagnosed or untreated

Temporal action localization is an important yet challenging task in video understanding. Typically, such a task aims at inferring both the action category and localization of the start and end frame for each action instance in a long, untrimmed video.While most current models achieve good results by using pre-defined anchors and numerous actionness, such methods could be bothered with both large number of outputs and heavy tuning of locations and sizes corresponding to different anchors. Instead, anchor-free methods is lighter, getting rid of redundant hyper-parameters, but gains few attention. In this paper, we propose the first purely anchor-free temporal localization method, which is both efficient and effective. Our model includes (i) an end-to-end trainable basic predictor, (ii) a saliency-based refinement module to gather more valuable boundary features for each proposal with a novel boundary pooling, and (iii) several consistency constraints to make sure our model can find the accurate boundary given arbitrary proposals. Extensive experiments show that our method beats all anchor-based and actionness-guided methods with a remarkable margin on THUMOS14, achieving state-of-the-art results, and comparable ones on ActivityNet v1.3. Code is available at https://github.com/TencentYoutuResearch/ActionDetection-AFSD., Accepted by CVPR2021

Super-Resolution (SR) is a fundamental computer vision task that aims to obtain a high-resolution clean image from the given low-resolution counterpart. This paper reviews the NTIRE 2021 Challenge on Video Super-Resolution. We present evaluation results from two competition tracks as well as the proposed solutions. Track 1 aims to develop conventional video SR methods focusing on the restoration quality. Track 2 assumes a more challenging environment with lower frame rates, casting spatio-temporal SR problem. In each competition, 247 and 223 participants have registered, respectively. During the final testing phase, 14 teams competed in each track to achieve state-of-the-art performance on video SR tasks.

For action recognition learning, 2D CNN-based methods are efficient but may yield redundant features due to applying the same 2D convolution kernel to each frame. Recent efforts attempt to capture motion information by establishing inter-frame connections while still suffering the limited temporal receptive field or high latency. Moreover, the feature enhancement is often only performed by channel or space dimension in action recognition. To address these issues, we first devise a Channel-wise Motion Enhancement (CME) module to adaptively emphasize the channels related to dynamic information with a channel-wise gate vector. The channel gates generated by CME incorporate the information from all the other frames in the video. We further propose a Spatial-wise Motion Enhancement (SME) module to focus on the regions with the critical target in motion, according to the point-to-point similarity between adjacent feature maps. The intuition is that the change of background is typically slower than the motion area. Both CME and SME have clear physical meaning in capturing action clues. By integrating the two modules into the off-the-shelf 2D network, we finally obtain a Comprehensive Motion Representation (CMR) learning method for action recognition, which achieves competitive performance on Something-Something V1 & V2 and Kinetics-400. On the temporal reasoning datasets Something-Something V1 and V2, our method outperforms the current state-of-the-art by 2.3% and 1.9% when using 16 frames as input, respectively., Accepted by AAAI21

Aiming at the problem that the previous tubular target leak detection method is not sensitive to small leakage and slow leakage and false alarm caused by external interference is not strong, tubular target leakage detection method based on deep learning is proposed. Firstly, the target detection network YOLO3 model is used to detect the tubular target in optical images. By analyzing the test results, the YOLO3 based tubular target leakage detection network is optimized and improved from three aspects: data set expansion, detection mode and network structure. Including: 1) using data transformation using rotation transformation, color dithering, zoom transformation, shift transformation and flip transformation on the data set; 2) according to the characteristics of the tubular target images, the detection method of polygon frame selection is used; 3) simplifying the network structure of the detection and output part. Finally, the improved network is trained and verified. The experimental results show that compared with the YOLO3 network model, the recognition accuracy and recall rate of the tubular target and the leaked area are greatly improved, and the average detection time is also reduced.

Deep learning methods have been more and more widely applied in the field of target detection. As an important part of deep learning target detection, non-maximum suppression is used to eliminate redundant detection bounding boxes generated during target detection and find out the optimal target boundary boxes, so as to speed up detection efficiency and improve detection accuracy. This article first introduces the related concepts and computational principles of traditional non-maximum suppression algorithm, and points out its problems. Based on this, the Soft-NMS, Softer-NMS, IOU-Guided NMS, Adaptive NMS and DIOU-NMS, a total of 5 kinds of improved maximum suppression algorithm principle is introduced and comparative analysis. And then we summarize the advantages and disadvantages of various algorithms. Finally, in view of the common problems existing in each algorithm, this paper points out the direction for improvement of non- maximum suppression algorithm, and provides technical reference and support for researchers in related fields.

Unsupervised learning of optical flow, which leverages the supervision from view synthesis, has emerged as a promising alternative to supervised methods. However, the objective of unsupervised learning is likely to be unreliable in challenging scenes. In this work, we present a framework to use more reliable supervision from transformations. It simply twists the general unsupervised learning pipeline by running another forward pass with transformed data from augmentation, along with using transformed predictions of original data as the self-supervision signal. Besides, we further introduce a lightweight network with multiple frames by a highly-shared flow decoder. Our method consistently gets a leap of performance on several benchmarks with the best accuracy among deep unsupervised methods. Also, our method achieves competitive results to recent fully supervised methods while with much fewer parameters., Accepted to CVPR 2020, https://github.com/lliuz/ARFlow

Motivated by the previous success of Two-Dimensional Convolutional Neural Network (2D CNN) on image recognition, researchers endeavor to leverage it to characterize videos. However, one limitation of applying 2D CNN to analyze videos is that different frames of a video share the same 2D CNN kernels, which may result in repeated and redundant information utilization, especially in the spatial semantics extraction process, hence neglecting the critical variations among frames. In this paper, we attempt to tackle this issue through two ways. 1) Design a sequential channel filtering mechanism, i.e., Progressive Enhancement Module (PEM), to excite the discriminative channels of features from different frames step by step, and thus avoid repeated information extraction. 2) Create a Temporal Diversity Loss (TD Loss) to force the kernels to concentrate on and capture the variations among frames rather than the image regions with similar appearance. Our method is evaluated on benchmark temporal reasoning datasets Something-Something V1 and V2, and it achieves visible improvements over the best competitor by \(2.4\%\) and \(1.3\%\), respectively. Besides, performance improvements over the 2D-CNN-based state-of-the-arts on the large-scale dataset Kinetics are also witnessed.

A direct and convenient approach for the coupling of propargylic substrates with diphenylphosphine oxide in the presence of Tf2O and 2,6-lutidine has been developed. The method provides a general a...

Most existing matching algorithms are one-off algorithms, i.e., they usually measure the distance between the two image feature representation vectors for only one time. In contrast, human's vision system achieves this task, i.e., image matching, by recursively looking at specific/related parts of both images and then making the final judgement. Towards this end, we propose a novel loopy recurrent neural network (Loopy RNN), which is capable of aggregating relationship information of two input images in a progressive/iterative manner and outputting the consolidated matching score in the final iteration. A Loopy RNN features two uniqueness. First, built on conventional long short-term memory (LSTM) nodes, it links the output gate of the tail node to the input gate of the head node, thus it brings up symmetry property required for matching. Second, a monotonous loss designed for the proposed network guarantees increasing confidence during the recursive matching process. Extensive experiments on several image matching benchmarks demonstrate the great potential of the proposed method., 6 pages, 5 figures, International Joint Conference on Artificial Intelligence(IJCAI)

When several high voltage direct current (HVDC) systems simultaneously recover after severe faults, absorbing large amounts of reactive power from the receiving AC systems, presents new challenges for the transient voltage stability of multi-infeed HVDC (MIDC) systems. Actually DC power recovery characteristics could be controllable to some extent by selecting voltage dependant current order limiter (VDCOL) parameters, but the quantitative relationship is complex and not easy to use. This paper, with experiment method based on PSCAD/EMTDC package and the CIGRE HVDC benchmark system, explored the converter bus voltage profile and the relationship of HVDC MW (P)and reactive power (Q) consumed by the converter during and after various faults. Through numerical analysis, the characteristic equation on P-Q has been fit out. With which and VDCOL characteristics, a simplified model of DC P-Q coupling recovery was constructed. Finally, based on the above model, a design method of VDCOL parameters was proposed with an objective function considering minimum sum of reactive power demand peaks of MIDC recovery. Some examples with different rated power combinations of MIDC systems were studied to demonstrate the feasibility of the proposed method.

The large demand of dynamic reactive power for MIDC recovery is a big challenge to voltage stability of the receiving system. Thus, the research on active power (P) and reactive power (Q) characteristics of inverter station in the recovery process of multi-infeed HVDC (MIDC) attracts more and more attention. With the CIGRE HVDC benchmark, this paper studies the P and Q responses under various faults through the electromagnetic transient simulation platform PSCAD/EMTDC. The major factors dominating P-Q behaviors are analyzed. Then, in order to obtain a relative simple and analytical relation between P (or Q) and HVDC control means in recovery for future theoretical stability analysis method, numerical analysis technique is used to establish reasonable fitting equations of V=f(V) and Q=f(P). When considering voltage dependant current order limiter (VDCOL) of HVDC, the VDCOL control equation and P-Q coupling equation is combined into a simplified model of DC recovery. It helps transient stability mechanism analysis of MIDC system in whole, only utilizing the external characteristic of HVDC and avoiding HVDC complex structure and control schemes. The obtained model of power recovery of HVDC is tested with good engineering accuracy shown.

Efficiency is an important issue in designing video architectures for action recognition. 3D CNNs have witnessed remarkable progress in action recognition from videos. However, compared with their 2D counterparts, 3D convolutions often introduce a large amount of parameters and cause high computational cost. To relieve this problem, we propose an efficient temporal module, termed as Temporal Enhancement-and-Interaction (TEI Module), which could be plugged into the existing 2D CNNs (denoted by TEINet). The TEI module presents a different paradigm to learn temporal features by decoupling the modeling of channel correlation and temporal interaction. First, it contains a Motion Enhanced Module (MEM) which is to enhance the motion-related features while suppress irrelevant information (e.g., background). Then, it introduces a Temporal Interaction Module (TIM) which supplements the temporal contextual information in a channel-wise manner. This two-stage modeling scheme is not only able to capture temporal structure flexibly and effectively, but also efficient for model inference. We conduct extensive experiments to verify the effectiveness of TEINet on several benchmarks (e.g., Something-Something V1&V2, Kinetics, UCF101 and HMDB51). Our proposed TEINet can achieve a good recognition accuracy on these datasets but still preserve a high efficiency., Accepted by AAAI 2020

Generating temporal action proposals remains a very challenging problem, where the main issue lies in predicting precise temporal proposal boundaries and reliable action confidence in long and untrimmed real-world videos. In this paper, we propose an efficient and unified framework to generate temporal action proposals named Dense Boundary Generator (DBG), which draws inspiration from boundary-sensitive methods and implements boundary classification and action completeness regression for densely distributed proposals. In particular, the DBG consists of two modules: Temporal boundary classification (TBC) and Action-aware completeness regression (ACR). The TBC aims to provide two temporal boundary confidence maps by low-level two-stream features, while the ACR is designed to generate an action completeness score map by high-level action-aware features. Moreover, we introduce a dual stream BaseNet (DSB) to encode RGB and optical flow information, which helps to capture discriminative boundary and actionness features. Extensive experiments on popular benchmarks ActivityNet-1.3 and THUMOS14 demonstrate the superiority of DBG over the state-of-the-art proposal generator (e.g., MGG and BMN). Our code will be made available upon publication., Accepted by AAAI 2020. Ranked No. 1 on ActivityNet Challenge 2019 on Temporal Action Proposals (http://activity-net.org/challenges/2019/evaluation.html)