375 research outputs found
Better Decremental and Fully Dynamic Sensitivity Oracles for Subgraph Connectivity
No full textWe study the sensitivity oracles problem for subgraph connectivity in the decremental and fully dynamic settings. In the fully dynamic setting, we preprocess an n-vertices m-edges undirected graph G with n_{off} deactivated vertices initially and the others are activated. Then we receive a single update D ⊆ V(G) of size |D| = d ≤ d_{⋆}, representing vertices whose states will be switched. Finally, we get a sequence of queries, each of which asks the connectivity of two given vertices u and v in the activated subgraph. The decremental setting is a special case when there is no deactivated vertex initially, and it is also known as the vertex-failure connectivity oracles problem.
We present a better deterministic vertex-failure connectivity oracle with Ô(d_{⋆}m) preprocessing time, Õ(m) space, Õ(d²) update time and O(d) query time, which improves the update time of the previous almost-optimal oracle [Long and Saranurak, 2022] from Ô(d²) to Õ(d²).
We also present a better deterministic fully dynamic sensitivity oracle for subgraph connectivity with Ô(min{m(n_{off} + d_{⋆}),n^{ω}}) preprocessing time, Õ(min{m(n_{off} + d_{⋆}),n²}) space, Õ(d²) update time and O(d) query time, which significantly improves the update time of the state of the art [Bingbing Hu et al., 2023] from Õ(d⁴) to Õ(d²). Furthermore, our solution is even almost-optimal assuming popular fine-grained complexity conjectures
Isolation and Impartial Aggregation: A Paradigm of Incremental Learning without Interference
No full textThis paper focuses on the prevalent stage interference and stage performance imbalance of incremental learning. To avoid obvious stage learning bottlenecks, we propose a new incremental learning framework, which leverages a series of stage-isolated classifiers to perform the learning task at each stage, without interference from others. To be concrete, to aggregate multiple stage classifiers as a uniform one impartially, we first introduce a temperature-controlled energy metric for indicating the confidence score levels of the stage classifiers. We then propose an anchor-based energy self-normalization strategy to ensure the stage classifiers work at the same energy level. Finally, we design a voting-based inference augmentation strategy for robust inference. The proposed method is rehearsal-free and can work for almost all incremental learning scenarios. We evaluate the proposed method on four large datasets. Extensive results demonstrate the superiority of the proposed method in setting up new state-of-the-art overall performance. Code is available at https://github.com/iamwangyabin/ESN
Learning Deep Trajectory Descriptor for action recognition in videos using deep neural networks
No full textHuman action recognition is widely recognized as a challenging task due to the difficulty of effectively characterizing human action in a complex scene. Recent studies have shown that the dense-trajectory-based methods can achieve state-of-the-art recognition results on some challenging datasets. However, in these methods, each dense trajectory is often represented as a vector of coordinates, consequently losing the structural relationship between different trajectories. To address the problem, this paper proposes a novel Deep Trajectory Descriptor (DTD) for action recognition. First, we extract dense trajectories from multiple consecutive frames and then project them onto a canvas. This will result in a 'trajectory texture' image which can effectively characterize the relative motion in these frames. Based on these trajectory texture images, a deep neural network (DNN) is utilized to learn a more compact and powerful representation of dense trajectories. In the action recognition system, the DTD descriptor, together with other non-trajectory features such as HOG, HOF and MBH, can provide an effective way to characterize human action from various aspects. Experimental results show that our system can statistically outperform several state-of-the-art approaches, with an average accuracy of 95:6% on KTH and an accuracy of 92.14% on UCF50. ? 2015 IEEE.EI2015-Augus
Modular graph attention network for complex visual relational reasoning
No full textVisual Relational Reasoning is crucial for many vision-and-language based tasks, such as Visual Question Answering and Vision Language Navigation. In this paper, we consider reasoning on complex referring expression comprehension (c-REF) task that seeks to localise the target objects in an image guided by complex queries. Such queries often contain complex logic and thus impose two key challenges for reasoning: (i) It can be very difficult to comprehend the query since it often refers to multiple objects and describes complex relationships among them. (ii) It is non-trivial to reason among multiple objects guided by the query and localise the target correctly. To address these challenges, we propose a novel Modular Graph Attention Network (MGA-Net). Specifically, to comprehend the long queries, we devise a language attention network to decompose them into four types: basic attributes, absolute location, visual relationship and relative locations, which mimics the human language understanding mechanism. Moreover, to capture the complex logic in a query, we construct a relational graph to represent the visual objects and their relationships, and propose a multi-step reasoning method to progressively understand the complex logic. Extensive experiments on CLEVR-Ref+, GQA and CLEVR-CoGenT datasets demonstrate the superior reasoning performance of our MGA-Net.Yihan Zheng, Zhiquan Wen, Mingkui Tan, Runhao Zeng, Qi Chen, Yaowei Wang, Qi W
Joint Learning of Semantic and Latent Attributes
No full textAs mid-level semantic properties shared across object categories, attributes have been studied extensively. Recent approaches have attempted joint modelling of multiple attributes together with class labels so as to exploit their correlations for better attribute prediction and object recognition. However, they often ignore the fact that there exist some shared properties other than nameable/semantic attributes, which we call latent attributes. Basically, they can be further divided into discriminative and non-discriminative parts depending on whether they can contribute to an object recognition task. We argue that learning the latent attributes jointly with user-defined semantic attributes not only leads to better representation for object recognition but also helps with semantic attribute prediction. A novel dictionary learning model is proposed which decomposes the dictionary space into three parts corresponding to semantic, latent discriminative and latent background attributes respectively. An efficient algorithm is then formulated to solve the resultant optimization problem. Extensive experiments show that the proposed attribute learning method produces state-of-the-art results on both attribute prediction and attribute-based person re-identification.CPCI-S(ISTP)[email protected]; [email protected]; [email protected]; [email protected]; [email protected]
ESUR: A SYSTEM FOR EVENTS DETECTION IN SURVEILLANCE VIDEO
No full textIn this paper, we present our eSur (Event detection system on SURveillance video) system, which is derived from TRECVID'09 surveillance tasks. Currently, eSur attempts to detect two categories of events: 1) single-actor events (i.e., PersonRuns and ElevatorNoEntry) irrespective of any interaction between individuals, and 2) pair-activity events (i. e., PeopleMeet, PeopleSplitUp, and Embrace) involves more than one individual. eSur consists of three major stages, i. e., preprocessing, event classification, and post-processing. The preprocessing involves view classification, background subtraction, head-shoulder detection, human body detection and object tracking. Event classification fuses One-vs.-All SVM and rule-based classifiers to identify single-actor and pair-activity events in an ensemble way. To reduce false alarms, we introduce prior knowledge into the post-processing, and in particular, we apply a so-called event merging process over TRECVID dataset. Extensive experiments have been performed over TRECVid'08 and '09 ED data corpus involving in total 144 hours surveillance video of London Gatwick airport. According to the TRECVid-ED formal evaluation, our prototype has yielded fairly promising results over TRECVid'09 dataset, with top Act. DCR of 1.023, 1.025, 1.02, and 0.334 for PeopleMeet, PeopleSplitUp, Embrace, and ElevatorNoEntry, respectively.http://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcApp=PARTNER_APP&SrcAuth=LinksAMR&KeyUT=WOS:000287728002101&DestLinkType=FullRecord&DestApp=ALL_WOS&UsrCustomerID=8e1609b174ce4e31116a60747a720701Engineering, Electrical & ElectronicImaging Science & Photographic TechnologyEICPCI-S(ISTP)
High-efficiency Coding for Shaking Surveillance Videos Based on Global Motion Compensation
No full textDue to the complex environment conditions, many surveillance videos are captured from cameras which are influenced by shaking more or less. This presents a significant challenge for background-modeling-based video coding since it is difficult to generate good background frames from such shaking videos. To solve this problem, this paper proposes a global motion compensation method using motion vectors (MV-GMC) for shaking surveillance video coding. In the proposed MV-GMC method, more accurate motion vectors (MVs) are extracted from HEVC encoder to estimate the global motion model in an efficient way, and we compensate each frame before background modeling. Then the compensated frames are used to model a good background frame for surveillance video coding. Compared with the optical-flow-based GMC (OPT-GMC) method which can be used to obtain more precise motion compensation, the proposed MV-GMC method has a comparable coding performance but a much lower computational complexity. Experiments on our surveillance video sequences show that the proposed MV-GMC method has significantly improved the coding performance by decreasing BD rate 49.83% over HM 12.0 on average while OPT-GMC can save 49.84% BD rate. The MVGMC method also saves 92.71% background modeling time compared with the OPT-GMC method.EICPCI-S(ISTP)[email protected]; [email protected]; [email protected]; [email protected]; [email protected]
CNN vs. SIFT for Image Retrieval: Alternative or Complementary?
No full textIn the past decade, SIFT is widely used in most vision tasks such as image retrieval. While in recent several years, deep convolutional neural networks (CNN) features achieve the state-of-the-art performance in several tasks such as image classification and object detection. Thus a natural question arises: for the image retrieval task, can CNN features substitute for SIFT? In this paper, we experimentally demonstrate that the two kinds of features are highly complementary. Following this fact, we propose an image representation model, complementary CNN and SIFT (CCS), to fuse CNN and SIFT in a multi-level and complementary way. In particular, it can be used to simultaneously describe scene level, object-level and point-level contents in images. Extensive experiments are conducted on four image retrieval benchmarks, and the experimental results show that our CCS achieves state-of-the-art retrieval results.CPCI-S(ISTP)[email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Impact of grazing on shaping abundance and composition of active methanotrophs and methane oxidation activity in a grassland soil
No full textThe effect of grazing on the abundance, composition, and methane (CH4) uptake of methanotrophs in grasslands has been well documented in the past few decades, but the dominant communities of active methanotrophs responsible for CH4 oxidation activity in grazed soils are still poorly understood. In this study, we characterized the metabolically active, aerobic methanotrophs in grasslands with three different levels of grazing (light, medium, and heavy) by combining DNA-stable isotope probing (SIP) and quantitative PCR (qPCR) for methane monooxygenase (pmoA) gene– and 16S rRNA gene–based amplicon sequencing. The CH4 oxidation potential was as low as 0.51 μmol g dry weight−1 day−1 in the ungrazed control, while it decreased as grazing intensity increased in grazed fields, ranging from 2.25 μmol g dry weight−1 day−1 in light grazed fields to 1.59 in heavily grazed fields. Increased CH4 oxidation activity was paralleled by twofold increases in abundance of pmoA genes and relative abundance of methanotroph-affiliated 16S rRNA genes in the total microbial community in grazed soils. SIP and sequencing revealed that the genera Methylobacter and Methylosarcina (type I; Gammaproteobacteria) and Methylocystis (type II; Alphaproteobacteria) were active methanotrophs responsible for CH4 oxidation in grazed soils. Light and intermediate grazing stimulated the growth and activity of methanotrophs, while heavy grazing decreased the abundance and diversity of the active methanotrophs in the typical steppe. Redundancy and correlation analysis further indicated that the variation of bulk density and soil C and N induced by grazing determined the abundance, diversity of active methanotrophs, and methane oxidation activity in the long-term grazed grassland soil
Atmospheric methane oxidation is affected by grassland type and grazing and negatively correlated to total soil respiration in arid and semiarid grasslands in Inner Mongolia
No full textMethane (CH
4) is an important trace greenhouse gas and atmospheric CH
4 uptake by high-affinity methanotrophs in grassland soil accounts for an important proportion of the terrestrial CH
4 sink. However, our understanding of the comprehensive effects of grassland type and grazing treatment on active soil methanotrophs and atmospheric CH
4 uptake is still under debate. This study investigates the impact of grazing on CH
4 oxidation rate and active atmospheric CH
4 oxidizing methanotroph communities in two arid and semiarid grassland ecosystems (meadow and desert) by detecting transcripts of methane monooxygenase (pmoA) genes. Atmospheric CH
4 oxidation rates differed according to grassland type and grazing treatment. The highest activity was found in desert grasslands with moderate grazing and the lowest activity in meadow grasslands with exclosures. The differences in activities were linked with changes in abundance, composition and co-occurrence network patterns of active methanotrophs and CO
2 production rate. Redundancy, correlation and random forest analyses indicated that pmoA transcripts, available phosphorus (AP), NO
3
−-N, and CO
2 production rate were the most important factors predicting active methanotroph community composition and atmospheric CH
4 oxidation activity in these grassland ecosystems. A glucose amendment incubation experiment showed that addition of glucose increased heterotrophic microbial respiration and inhibited atmospheric CH
4 oxidation. This study provides evidence that CO
2 production rate is an important factor associated with atmospheric CH
4 oxidation activity in arid and semiarid grassland ecosystems and suggests that interactions between methanotrophs and other heterotrophs influence methanotroph activity in grassland ecosystems.
</p
- …
