scholarly journals Multimodal Transformer Networks for Pedestrian Trajectory Prediction

2021 ◽  
Author(s):  
Ziyi Yin ◽  
Ruijin Liu ◽  
Zhiliang Xiong ◽  
Zejian Yuan
Keyword(s):  
Optical Flow ◽  
Temporal Variations ◽  
Coarse Grained ◽  
Vehicle Speed ◽  
Temporal Consistency ◽  
Moving Vehicle ◽  
Dynamic Motion ◽  
Fine Grained ◽  
High Dynamics

We consider the problem of forecasting the future locations of pedestrians in an ego-centric view of a moving vehicle. Current CNNs or RNNs are flawed in capturing the high dynamics of motion between pedestrians and the ego-vehicle, and suffer from the massive parameter usages due to the inefficiency of learning long-term temporal dependencies. To address these issues, we propose an efficient multimodal transformer network that aggregates the trajectory and ego-vehicle speed variations at a coarse granularity and interacts with the optical flow in a fine-grained level to fill the vacancy of highly dynamic motion. Specifically, a coarse-grained fusion stage fuses the information between trajectory and ego-vehicle speed modalities to capture the general temporal consistency. Meanwhile, a fine-grained fusion stage merges the optical flow in the center area and pedestrian area, which compensates the highly dynamic motion of ego-vehicle and target pedestrian. Besides, the whole network is only attention-based that can efficiently model long-term sequences for better capturing the temporal variations. Our multimodal transformer is validated on the PIE and JAAD datasets and achieves state-of-the-art performance with the most light-weight model size. The codes are available at https://github.com/ericyinyzy/MTN_trajectory.

Studies with the intraruminal selenium pellet. 2. The effect of grain size of selenium on the functional life of pellets in sheep

1981 ◽  
Vol 32 (6) ◽  
pp. 935 ◽  
Author(s):  
DR Hudson ◽  
RA Hunter ◽  
DW Peter
Keyword(s):  
Grain Size ◽  
Progressive Increase ◽  
Coarse Grained ◽  
Elemental Selenium ◽  
Average Composition ◽  
Plasma Selenium ◽  
Iron Selenide ◽  
Fine Grained ◽  
Grain Sizes

Grain size of elemental selenium is a major factor controlling the long-term effectiveness of intraruminal selenium pellets. Microscope studies of polished sections of new and used selenium pellets showed that two commercially manufactured pellets contained selenium with average grain sizes about 4 and 40 �m respectively. Plasma selenium concentrations in sheep treated with pellets containing the coarse-grained selenium were maintained at higher levels over longer periods of time than those measured for sheep treated with pellets with fine-grained selenium. Pellets removed from sheep after 2, 4, 8, 16 and 28 days showed a progressive increase in the degree of alteration of selenium to a compound of average composition (g/100 g) iron, 33.7; selenium, 51.3 ; oxygen, 15.0. After 28 days only a small percentage of elemental selenium remained in pellets with fine-grained selenium, whereas about 50% remained in pellets with coarse-grained selenium. CSIRO prototype pellets, for which long-term effectiveness had been established, also contained coarse-grained selenium, and remnants of selenium were found in pellets that had been in sheep for periods up to 3 years. Selenium, administered in gelatin capsules or as sachets containing glass-selenium mixtures, was stable under the pH-Eh conditions of the rumen, but was rendered unstable in selenium pellets or iron-selenium mixtures by the presence of iron. It is probable that the most rapid release of selenium to the sheep occurs as a result of a chemical reaction involving the oxidation of iron and concomitant alteration of elemental selenium to iron selenide.

Deformation and reaction fabrics in eclogites from the Western Gneiss Region (Norway) - evidence of dehydration reactions attributed to episodic deformation

2021 ◽  
Author(s):  
Claudia A. Trepmann ◽  
Ane K. Engvik ◽  
Erick G. Prince Gutierrez
Keyword(s):  
Inner Core ◽  
White Mica ◽  
Coarse Grained ◽  
Dislocation Creep ◽  
Western Gneiss Region ◽  
Garnet Porphyroblasts ◽  
Boundary Area ◽  
Fine Grained ◽  
Dehydration Reactions

<p>The eclogites from Vårdalneset, Western Gneiss Region, Norway, show an exceptional large variety of reaction and deformation microfabrics that document the processes and conditions during burial and exhumation. Coarse grained eclogites comprise about 35% omphacite, 25% garnet and 20% amphibole with various amounts of white mica, zoisite, kyanite, rutile, zircon and pyrite. Their fabric is characterized by few mm long and several hundred µm wide amphibole and omphacite grains aligned in the foliation plane with zoned garnet porphyroblasts up to several mm in diameter. In contrast, finer-grained mylonitic eclogites with grain diameters of few hundred µm comprise systematically higher amounts of garnet (45%) and omphacite (35%) and generally less amphibole (< 5%) but similar amounts of zoisite, white mica, rutile and quartz. In the coarse-grained eclogite, amphibole shows evidence of dislocation creep as indicated by undulatory extinction, subgrains and recrystallized grains in necks of boudinaged coarse amphibole layers as well as in contact to garnet. The large garnet porphyroblasts generally show a complex zonation with an inclusion-rich Fe-poor and Mg-rich inner core surrounded by a zone with Fe- and Ca-rich patches and a broad Mg-rich, Ca- and Fe-poor rim. Only at contact to coarse amphibole an additional, a few tens of µm thin serrated rim further enriched in Mg can occur. At the direct contact to such serrated Mg-rich rims, amphibole is partly replaced by a fine-grained quartz-kyanite ± rutile aggregate, indicating dehydration reactions of amphibole. Quartz - kyanite ± rutile aggregates are surrounding garnet also in contact to omphacite, zoisite and to other garnet crystals. The microstructures suggest that deformation and dehydration of amphibole are coupled and played an important role during deformation of the eclogites finally leading to the mylonitic eclogites with higher amounts of garnet and omphacite. Deformation is suggested to have triggered the dehydration reaction by a slight and local increase in temperature. Furthermore, deformation provided additional pathways for the escaping fluids along the increased grain and phase boundary area, as indicated by commonly present quartz within interstitials between recrystallized amphibole grains. In all samples, few µm wide amphibole rims replacing garnets document restricted rehydration-reactions at a later stage. The large variety of the deformation and reaction microfabrics exemplarily show that both deformation and metamorphic reactions did not proceed at long-term continuous conditions, but that both are coupled and occurred episodically.</p>

scholarly journals Hydrogeological dissimilarity of geodynamically different terranes

2021 ◽  
Vol 44 (2) ◽  
pp. 134-140
Author(s):  
V. E. Glotov
Keyword(s):  
Continental Margin ◽  
Sandy Loam ◽  
Active Continental Margin ◽  
Passive Continental Margin ◽  
Coarse Grained ◽  
Active Margin ◽  
Water Abundance ◽  
Fine Grained ◽  
Weathering Processes

The article presents and analyzes the data on ground waters of active (suprapermafrost) and hindered (subpermafrost) water exchange of geodynamically different terrains in order to prove the hydrogeological importance of their historical and tectonic characteristics. On the example of Trans-Polar Chukotka it is shown that, under suprapermafrost conditions, the ubiquitous eluvial-deluvial nappes are the most water-abundant on the terrane – a fragment of the passive continental margin, whereas they are the least water-abundant on the terrains of the active margin. Hydrogeological situation changes under subpermafrost conditions: more permeable and water-retaining rocks compose the terranes of the active margin. These differences are associated with the level of rock tectonic decompaction and, accordingly, with different intensity of weathering processes in the terrane rocks of different geodynamic origin in suprapermafrost and subpermafrost conditions. The hypergenesis zone on the terranes of the passive continental margin features coarse-grained rock weathering products accumulated in relatively calm geological and historical environments, the aggregate is sandy. The terranes of the active margin, which underwent long-term subvertical and subhorizontal displacements contain more fine-grained weathering products; the aggregate includes sandy loam and clay sand. Since the permafrost strata in both Trans-Polar Chukotka and Eastern Siberia is greater than the depth of hypergene transformations, the terranes of the active continental margin, the rocks of which were impacted by tectonic decompaction processes, mainly of a strike-slip and thrust nature, feature greater water abundance in subpermafrost conditions.

Imposed strain localization in the mantle section of an oceanic transform zone revealed by microstructural and stress variations

2020 ◽  
Author(s):  
Vasileios Chatzaras ◽  
Basil Tikoff ◽  
Seth C. Kruckenberg ◽  
Sarah J. Titus ◽  
Christian Teyssier ◽  
...  
Keyword(s):  
Shear Zone ◽  
Upper Mantle ◽  
Temporal Variations ◽  
Coarse Grained ◽  
Transform Fault ◽  
Differential Stress ◽  
Transform Faults ◽  
Fine Grained ◽  
Micro Deformation ◽  
Mantle Section

<p>Mantle earthquakes that occur deeper than the 600 °C isotherm in oceanic transform faults indicate seismic rupturing at conditions where viscous deformation (bulk ductile behavior) is dominant.  However, direct geological evidence of earthquake-related deformation at ambient upper mantle conditions is rare, impeding our understanding of earthquake dynamics in plate-boundary fault systems.  The Bogota Peninsula Shear Zone (BPSZ), New Caledonia, is an ancient oceanic transform fault exhumed from upper mantle depths.  Ductile structures in the BPSZ formed at temperatures > 800 °C and microstructures indicate that differential stress varies spatially and temporally.  Spatial variation is observed as an increase in differential stress with strain toward localized zones of high strain; stress increases from 6–14 MPa in coarse grained tectonites to 11–22 MPa within 1–2 km wide mylonite zones.  Temporal stress variation is observed by the formation of micro-deformation zones that seem to have brittle precursors, are filled with fine-grained recrystallized olivine grains and crosscut the background fabrics in the harzburgites that host them.  The micro-deformation zones are not restricted to the mylonite zones, but rather are located throughout the BPSZ, having affected the protomylonites and the coarse grained tectonites.  The micro-deformation zones record stresses of 22–81 MPa that are 2–6 times higher than the background, steady-state stresses in the surrounding mantle rocks.  We interpret the observed spatial and temporal variations in microstructures and stresses in the upper mantle to demonstrate the influence of seismic events in the upper part of the oceanic transform fault system.  We attribute the increase in stress with strain to be the result of imposed localization induced by downward propagation of the seismic rupture into the underlying mantle.  The micro-deformation zones could result from brittle fractures caused by earthquake-related deformation in the mantle section of the transform fault, which are in turn overprinted by ductile deformation.</p><p> </p><p>Synthesizing the spatial and temporal variations in stresses and microstructures in the Bogota Peninsula Shear Zone we propose a conceptual model where brittle fracturing and shearing take place during coseismic rupture at increased stress, ductile flow at decaying stress is concentrated in the micro-deformation zones during postseismic relaxation, and uniformly distributed creep at low stress occurs in the host-rocks of the micro-deformation zones during interseismic deformation.  The critical result from the studied paleotransform zone is that the fine-grained micro-deformation zones and the mylonites do not represent weak zones.  Instead, they form by dislocation creep at transient high-stress deformation during the seismic cycle.  The spatial distribution of the micro-deformation zones also suggests that repeated stress cycles in oceanic transform faults may not localize strain in pre-existing shear zones but disperse strain across the structure.</p>

Correlation between dislocation, grain boundary and interface of duplex SS in stress corrosion cracking(SCC)

1990 ◽  
Vol 48 (4) ◽  
pp. 928-929
Author(s):  
Wang Zheng-fang ◽  
Z.F. Wang
Keyword(s):  
Stainless Steel ◽  
Thermal Cycle ◽  
Secondary Phase ◽  
Corrosion Cracking ◽  
Coarse Grained ◽  
Test Environment ◽  
Fine Grained ◽  
Evaluation Test ◽  
Welding Thermal Cycle ◽  
Micro Analysis

The main purpose of this study highlights on the evaluation of chloride SCC resistance of the material,duplex stainless steel,OOCr18Ni5Mo3Si2 (18-5Mo) and its welded coarse grained zone(CGZ).18-5Mo is a dual phases (A+F) stainless steel with yield strength:512N/mm2 .The proportion of secondary Phase(A phase) accounts for 30-35% of the total with fine grained and homogeneously distributed A and F phases(Fig.1).After being welded by a specific welding thermal cycle to the material,i.e. Tmax=1350°C and t8/5=20s,microstructure may change from fine grained morphology to coarse grained morphology and from homogeneously distributed of A phase to a concentration of A phase(Fig.2).Meanwhile,the proportion of A phase reduced from 35% to 5-10°o.For this reason it is known as welded coarse grained zone(CGZ).In association with difference of microstructure between base metal and welded CGZ,so chloride SCC resistance also differ from each other.Test procedures:Constant load tensile test(CLTT) were performed for recording Esce-t curve by which corrosion cracking growth can be described, tf,fractured time,can also be recorded by the test which is taken as a electrochemical behavior and mechanical property for SCC resistance evaluation. Test environment:143°C boiling 42%MgCl2 solution is used.Besides, micro analysis were conducted with light microscopy(LM),SEM,TEM,and Auger energy spectrum(AES) so as to reveal the correlation between the data generated by the CLTT results and micro analysis.

scholarly journals Domain randomization-enhanced deep learning models for bird detection

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Xin Mao ◽  
Jun Kang Chow ◽  
Pin Siang Tan ◽  
Kuan-fu Liu ◽  
Jimmy Wu ◽  
...  
Keyword(s):  
Deep Learning ◽  
Continuous Monitoring ◽  
Bird Species ◽  
Training Data ◽  
Learning Models ◽  
Fine Grained ◽  
Bird Detection ◽  
Relationship Of ◽  
The Relationship

AbstractAutomatic bird detection in ornithological analyses is limited by the accuracy of existing models, due to the lack of training data and the difficulties in extracting the fine-grained features required to distinguish bird species. Here we apply the domain randomization strategy to enhance the accuracy of the deep learning models in bird detection. Trained with virtual birds of sufficient variations in different environments, the model tends to focus on the fine-grained features of birds and achieves higher accuracies. Based on the 100 terabytes of 2-month continuous monitoring data of egrets, our results cover the findings using conventional manual observations, e.g., vertical stratification of egrets according to body size, and also open up opportunities of long-term bird surveys requiring intensive monitoring that is impractical using conventional methods, e.g., the weather influences on egrets, and the relationship of the migration schedules between the great egrets and little egrets.

scholarly journals Surface Water Temperature Predictions at a Mid-Latitude Reservoir under Long-Term Climate Change Impacts Using a Deep Neural Network Coupled with a Transfer Learning Approach

Water ◽  
2021 ◽  
Vol 13 (8) ◽  
pp. 1109
Author(s):  
Nobuaki Kimura ◽  
Kei Ishida ◽  
Daichi Baba
Keyword(s):  
Climate Change ◽  
Surface Water ◽  
Water Temperature ◽  
Transfer Learning ◽  
Air Temperature ◽  
Climate Models ◽  
Temporal Variations ◽  
Training Data ◽  
Surface Water Temperature

Long-term climate change may strongly affect the aquatic environment in mid-latitude water resources. In particular, it can be demonstrated that temporal variations in surface water temperature in a reservoir have strong responses to air temperature. We adopted deep neural networks (DNNs) to understand the long-term relationships between air temperature and surface water temperature, because DNNs can easily deal with nonlinear data, including uncertainties, that are obtained in complicated climate and aquatic systems. In general, DNNs cannot appropriately predict unexperienced data (i.e., out-of-range training data), such as future water temperature. To improve this limitation, our idea is to introduce a transfer learning (TL) approach. The observed data were used to train a DNN-based model. Continuous data (i.e., air temperature) ranging over 150 years to pre-training to climate change, which were obtained from climate models and include a downscaling model, were used to predict past and future surface water temperatures in the reservoir. The results showed that the DNN-based model with the TL approach was able to approximately predict based on the difference between past and future air temperatures. The model suggested that the occurrences in the highest water temperature increased, and the occurrences in the lowest water temperature decreased in the future predictions.

scholarly journals Deep Learning-Based Congestion Detection at Urban Intersections

Sensors ◽  
2021 ◽  
Vol 21 (6) ◽  
pp. 2052
Author(s):  
Xinghai Yang ◽  
Fengjiao Wang ◽  
Zhiquan Bai ◽  
Feifei Xun ◽  
Yulin Zhang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Optical Flow ◽  
Traffic Congestion ◽  
Detection Algorithm ◽  
Input Image ◽  
Vehicle Speed ◽  
Position Information ◽  
Traffic State ◽  
State Discrimination ◽  
Discrimination Method

In this paper, a deep learning-based traffic state discrimination method is proposed to detect traffic congestion at urban intersections. The detection algorithm includes two parts, global speed detection and a traffic state discrimination algorithm. Firstly, the region of interest (ROI) is selected as the road intersection from the input image of the You Only Look Once (YOLO) v3 object detection algorithm for vehicle target detection. The Lucas-Kanade (LK) optical flow method is employed to calculate the vehicle speed. Then, the corresponding intersection state can be obtained based on the vehicle speed and the discrimination algorithm. The detection of the vehicle takes the position information obtained by YOLOv3 as the input of the LK optical flow algorithm and forms an optical flow vector to complete the vehicle speed detection. Experimental results show that the detection algorithm can detect the vehicle speed and traffic state discrimination method can judge the traffic state accurately, which has a strong anti-interference ability and meets the practical application requirements.

scholarly journals Non-Local and Multi-Scale Mechanisms for Image Inpainting

Sensors ◽  
2021 ◽  
Vol 21 (9) ◽  
pp. 3281
Author(s):  
Xu He ◽  
Yong Yin
Keyword(s):  
Markov Random Fields ◽  
Receptive Fields ◽  
Image Inpainting ◽  
Long Distance ◽  
Visual Appearance ◽  
Fine Grained ◽  
Multi Scale ◽  
Non Local ◽  
Relationship Of

Recently, deep learning-based techniques have shown great power in image inpainting especially dealing with squared holes. However, they fail to generate plausible results inside the missing regions for irregular and large holes as there is a lack of understanding between missing regions and existing counterparts. To overcome this limitation, we combine two non-local mechanisms including a contextual attention module (CAM) and an implicit diversified Markov random fields (ID-MRF) loss with a multi-scale architecture which uses several dense fusion blocks (DFB) based on the dense combination of dilated convolution to guide the generative network to restore discontinuous and continuous large masked areas. To prevent color discrepancies and grid-like artifacts, we apply the ID-MRF loss to improve the visual appearance by comparing similarities of long-distance feature patches. To further capture the long-term relationship of different regions in large missing regions, we introduce the CAM. Although CAM has the ability to create plausible results via reconstructing refined features, it depends on initial predicted results. Hence, we employ the DFB to obtain larger and more effective receptive fields, which benefits to predict more precise and fine-grained information for CAM. Extensive experiments on two widely-used datasets demonstrate that our proposed framework significantly outperforms the state-of-the-art approaches both in quantity and quality.

Sign in / Sign up

Export Citation Format

Share Document