In this work, we introduce a novel task - Human-centric Spatio-Temporal Video Grounding (HC-STVG). Unlike the existing referring expression tasks in images or videos, by focusing on humans, HC-STVG ...aims to localize a spatio-temporal tube of the target person from an untrimmed video based on a given textural description. This task is useful, especially for healthcare and security related applications, where the surveillance videos can be extremely long but only a specific person during a specific period is concerned. HC-STVG is a video grounding task that requires both spatial (where) and temporal (when) localization. Unfortunately, the existing grounding methods cannot handle this task well. We tackle this task by proposing an effective baseline method named Spatio-Temporal Grounding with Visual Transformers (STGVT), which utilizes Visual Transformers to extract cross-modal representations for video-sentence matching and temporal localization. To facilitate this task, we also contribute an HC-STVG datasetThe new dataset is available at https://github.com/tzhhhh123/HC-STVG . consisting of 5,660 video-sentence pairs on complex multi-person scenes. Specifically, each video lasts for 20 seconds, pairing with a natural query sentence with an average of 17.25 words. Extensive experiments are conducted on this dataset, demonstrating that the newly-proposed method outperforms the existing baseline methods.
Weakly-supervised phrase grounding aims to localize a specific region in an image that corresponds to the given textual phrase, where the mapping between noun phrases and image regions is not ...available in the training stage. Previous methods typically exploit an additional proxy task (e.g., phrase reconstruction or image-phrase alignment) to provide supervision for training, since the lack of region-level annotations in the weakly-supervised setting. However, there exists a significant gap in optimization objectives between the proxy tasks and the target grounding task, which may result in low-efficient optimization for the target model. Therefore, in this paper, we propose a novel dual reinforcement learning framework to directly optimize the phrase grounding model. Specifically, we consider the duality of phrase grounding and phrase generation tasks. These two tasks form a closed loop that can provide quality feedback signals to measure the performance of each other. In this way, we can measure the correctness of the localized regions and thus be able to optimize the grounding model directly. We design two reward functions to quantify the feedback signals and train the models via reinforcement learning. In addition, to relieve the training difficulty of our framework, we present a heuristic algorithm to generate pseudo region-phrase pairs to warm-start our models. We perform experiments on two popular phrase grounding datasets: ReferItGame and Flickr30 K Entities, and the results demonstrate that our method outperforms the previous methods by a large margin.
Grounding-zone wedges (GZWs) are asymmetric sedimentary depocentres which form through the rapid accumulation of glacigenic debris along a line-source at the grounding zone of marine-terminating ice ...sheets during still-stands in ice-sheet retreat. GZWs form largely through the delivery of deforming subglacial sediments. The presence of GZWs in the geological record indicates an episodic style of ice retreat punctuated by still-stands in grounding-zone position. Moraine ridges and ice-proximal fans may also build up at the grounding zone during still-stands of the ice margin, but these require either considerable vertical accommodation space or sediment derived from point-sourced subglacial meltwater streams. By contrast, GZWs form mainly where floating ice shelves constrain vertical accommodation space immediately beyond the grounding-zone. An inventory of GZWs is compiled from available studies of bathymetric and acoustic data from high-latitude continental margins. The locations and dimensions of GZWs from the Arctic and Antarctic, alongside a synthesis of their key architectural and geomorphic characteristics, are presented. GZWs are only observed within cross-shelf troughs and major fjord systems, which are the former locations of ice streams and fast-flowing outlet glaciers. Typical high-latitude GZWs are less than 15km in along-flow direction and 15 to 100m thick. GZWs possess a transparent to chaotic acoustic character, which reflects the delivery of diamictic subglacial debris. Many GZWs contain seaward-dipping reflections, which indicate sediment progradation and wedge-growth through continued delivery of basal sediments. GZW formation is inferred to require high rates of sediment delivery to a fast-flowing ice margin that is relatively stable for probably decades to centuries. Although the long-term stability of the grounding zone is controlled by ice-sheet mass balance, the precise location of any still-stands is influenced strongly by the geometry of the continental shelf. The majority of high-latitude GZWs occur at vertical or lateral pinning points, which encourage grounding-zone stabilisation through increasing basal and lateral drag and reducing mass flow across the grounding zone.
•We present an inventory of 143 high-latitude grounding-zone wedges.•Grounding-zone wedges are typically less than 15km long and 15 to 100m thick.•Fast ice-streaming flow is probably necessary for grounding-zone wedge formation.•Ice shelves force wedge formation by restricting vertical accommodation space.•Continental shelf geometry can control the position of still-stands in the ice margin.
A grounding grid is essential to the lightning protection of power systems. This paper presents a modified partial element equivalent circuit method for predicting the transient behavior of the ...grounding grid. The frequency-dependent parameters of the grounding grid are obtained first by using the image method. Both modified nodal formulation and vector fitting techniques are applied to derive an extended equivalent network for time-domain simulation. In this method, the soil ionization effect is considered using a nonlinear resistance. The proposed method is verified with experimental results available in the literature. Finally, lightning transients in the grounding grid of a radio base station is presented. The ionization and propagation effects on grounding grid performance are discussed.
In this paper we present how automatic maritime anomaly detection tools can be successfully applied in real-world situations such as the major event of the container vessel Ever Given , which ...grounded in the Suez Canal on March 23rd 2021. The anomaly detector is designed to process the available sequence of Automatic Identification System (AIS) reports, information from ground-based or satellite radar systems if available, and contextual information defining the expected nominal behavior of navigation. A statistical hypothesis testing procedure is sequentially run to decide whether or not a deviation from the nominal behavior happened within a specific time period, for instance two consecutive data points. We show, based on the recorded AIS data from the Ever Given , that the proposed detector could have been triggered and alerted to anomalous behavior fully 19 minutes before the grounding.
Weakly supervised temporal sentence grounding has better scalability and practicability than fully supervised methods in real-world application scenarios. However, most of existing methods cannot ...model the fine-grained video-text local correspondences well and do not have effective supervision information for correspondence learning, thus yielding unsatisfying performance. To address the above issues, we propose an end-to-end Local Correspondence Network (LCNet) for weakly supervised temporal sentence grounding. The proposed LCNet enjoys several merits. First, we represent video and text features in a hierarchical manner to model the fine-grained video-text correspondences. Second, we design a self-supervised cycle-consistent loss as a learning guidance for video and text matching. To the best of our knowledge, this is the first work to fully explore the fine-grained correspondences between video and text for temporal sentence grounding by using self-supervised learning. Extensive experimental results on two benchmark datasets demonstrate that the proposed LCNet significantly outperforms existing weakly supervised methods.
Hybrid Grounding of Electrical Systems Panetta, Sergio A. R.
IEEE transactions on industry applications,
2017-Nov.-Dec., 2017-11-00, Letnik:
53, Številka:
6
Journal Article
Recenzirano
This paper will discuss the many different ways to ground medium voltage generators. Single source as well as multiple source units, either unit connected or bus connected, will be shown with ...preferred grounding schemes.
Grounding system design ensures safety compliance for transmission lines under lightning conditions. In this article, the transient response of short concrete encased grounding electrodes is ...investigated, that is, a 2.5 m vertical rod and a 10 m horizontal electrode (HE) (counterpoise). Situations of first and subsequent lightning strokes applied to horizontal and vertical concrete encapsulated electrodes are analyzed. The effect of the frequency dependence of soil parameters is considered. Also, soils with low, medium, and high resistivities <inline-formula><tex-math notation="LaTeX">(\text{300}{\bm{ }}</tex-math></inline-formula> to <inline-formula><tex-math notation="LaTeX">\text{4000}{\bm{\ }}\Omega .m)</tex-math></inline-formula> are evaluated. Different lengths of concrete encasement are considered in order to verify situations with either partial or total encapsulation. For all these cases, the harmonic impedance (from 60 Hz to 4 MHz) is calculated, and the respective ground potential rise is derived for median lightning currents. The results obtained indicate that concrete encased electrodes produced reductions in the impulse impedance up to 30% for the HE and 44% for the vertical electrode. Moreover, for the evaluated situations, the percentage reduction on impulse impedance was only slightly affected by the increase in soil resistivity, which is also observed for soils modeled with constant parameters.
Human-centric instructional videos provide opportunities for users to learn real-world multi-step tasks, such as cooking, makeup, and using professional tools, etc. However, these lengthy videos ...always lead to a tedious learning experience, making it challenging for learners to catch specific guidance efficiently. In this paper, we present a novel approach, named Fuzzy Multimodal Graph Reasoning (FMGR) , to extract target events in long untrimmed human-centric instructional videos using natural language. Specifically, we devise a Fuzzy Multimodal Graph Learning layers in our method, which encompass (1) Contextual Graph Reasoning that transforms the individual features into contextualized features, (2) Cross-Modal Relation Fuzzifier that models the fine-grained matching relationships between two modalities, and (3) Fuzzy Graph Reasoning that conducts massage passing among cross-modal matching node pairs. Particularly, we integrate fuzzy theory into the cross-modal relation fuzzifier to amplify potential matching pairs, while simultaneously mitigating the interference from ambiguous matches. To validate our method, we conducted evaluations on two human-centric instructional video datasets, i.e. , MedVidQA and YouMakeUp. Moreover, we also take further analysis on the impacts of interrogative and declarative queries. Extensive experimental results and further analysis reveal the effectiveness of our proposed FMGR method.
PDF
