Understanding urban scenes is a fundamental ability requirement for assisted driving and autonomous vehicles. Most of the available scene understanding methods use red-green-blue (RGB) images; however, their segmentation performances are prone to degradation under adverse lighting conditions. Recently, many effective artificial neural networks have been presented shown that incorporating RGB thermal (RGB-T) images can improve accuracy even unsatisfactory However, potential multimodal feature...

RGB–thermal scene parsing has recently attracted increasing research interest in the field of computer vision. However, most existing methods fail to perform good boundary extraction for prediction maps and cannot fully use high-level features. In addition, these simply fuse features from RGB thermal modalities but are unable obtain comprehensive fused To address problems, we propose an edge-aware guidance fusion network (EGFNet) parsing. First, introduce a prior edge map generated using...

Color–thermal (RGB-T) urban scene parsing has recently attracted widespread interest. However, most existing approaches to RGB-T do not deeply explore the information complementarity between features. In this study, we propose a cross-modal attention-cascaded fusion network (CACFNet) that fully exploits cross-modality. our design, attention module mines complementary from two modalities. Subsequently, cascaded decodes multi-level features in an up-bottom manner. Noting each pixel is labeled...

Urban scene parsing is the core of intelligent transportation system, and RGB–thermal urban has recently attracted increasing research interest in field computer vision. However, most existing approaches fail to perform good boundary extraction for prediction maps cannot fully use high-level features. In addition, these methods simply fuse features from RGB thermal modalities but are unable obtain comprehensive fused To address problems, an edge-aware guidance fusion network (EGFNet) was...

In this paper, we introduce a novel benchmark, dubbed VastTrack, towards facilitating the development of more general visual tracking via encompassing abundant classes and videos. VastTrack possesses several attractive properties: (1) Vast Object Category. particular, it covers target objects from 2,115 classes, largely surpassing object categories existing popular benchmarks (e.g., GOT-10k with 563 LaSOT 70 categories). With such vast expect to learn tracking. (2) Larger scale. Compared...

Abstract Multimodal sentiment analysis is a downstream branch task of with high attention at present. Previous work in multimodal have focused on the representation and fusion modalities, capturing underlying semantic relationships between modalities by considering contextual information. While this approach feasible for simple comments, more complex comments require integration external knowledge to obtain accurate However, incorporating into enhance information complementarity has not been...

RGB-T (red–green–blue and thermal) scene parsing has recently drawn considerable research attention. Although existing methods efficiently conduct parsing, their performance remains limited by a small receptive field. Unlike that capture the global context fusing multiscale features or using an attention mechanism, we propose graph-enhancement branch network (GEBNet), which uses long-range dependencies obtained from to refine coarse semantic map generated decoder. Semantic detail modules...

High-performance Transformer trackers have shown excellent results, yet they often bear a heavy computational load. Observing that smaller input can immediately and conveniently reduce computations without changing the model, an easy solution is to adopt low-resolution for efficient tracking. Albeit faster, this hurts tracking accuracy much due information loss in low resolution In paper, we aim mitigate such boost performance of via dual knowledge distillation from frozen high-resolution...

Multimodal (e.g., RGB-Depth/RGB-Thermal) fusion has shown great potential for improving semantic segmentation in complex scenes indoor/low-light conditions). Existing approaches often fully fine-tune a dual-branch encoder-decoder framework with complicated feature strategy achieving multimodal segmentation, which is training-costly due to the massive parameter updates extraction and fusion. To address this issue, we propose surprisingly simple yet effective dual-prompt learning network...

Petrochemical equipment detection technology plays important role in petrochemical industry security monitoring systems, working status analysis and other applications. In complex scenes, the accuracy speed of would be limited because missing false with extreme sizes, due to image quality, scale, light, factors. this paper, a one-stage attention mechanism-enhanced Yolov5 network is proposed detect typical types scene images. The model considers advantages channel spatial mechanism...

RGB thermal scene parsing has recently attracted increasing research interest in the field of computer vision. However, most existing methods fail to perform good boundary extraction for prediction maps and cannot fully use high level features. In addition, these simply fuse features from modalities but are unable obtain comprehensive fused To address problems, we propose an edge-aware guidance fusion network (EGFNet) parsing. First, introduce a prior edge map generated using images capture...

Current multi-object tracking (MOT) aims to predict trajectories of targets (i.e.,"where") in videos. Yet, knowing merely "where" is insufficient many crucial applications. In comparison, semantic understanding such as fine-grained behaviors, interactions, and overall summarized captions (i.e., "what") from videos, associated with "where", highly-desired for comprehensive video analysis. Thus motivated, we introduce Semantic Multi-Object Tracking (SMOT), that estimate object meanwhile...

Abstract Non-destructive testing is a crucial process to ensure the quality of pipeline welding. Phased Array Ultrasonic Testing (PAUT) provides real-time scanning and imaging, compared radiogaphic testing, PAUT harmless humans environment, with accurate defect localization high sensitivity. However, due its relatively short history widespread adoption, there shortage professionals teams trained in traditional techniques such as radiography. In this paper, we study intelligent recognition...

Abstract Pipeline transportation serves as the primary method for conveying oil and gas, with welding being predominant means of connecting pipelines. Nevertheless, weld defects frequently occur at circumferential welds due to various factors including process environmental conditions. Failures these threaten safety pipeline transportation, X-ray inspection is used visually detect defects. While manual has traditionally been employed defect detection, it suffers from inefficiency...

Abstract In this paper, the deformation behaviour of large steel tank under uneven foundation settlement is investigated. The finite element model firstly established, analysis harmonic conditions carried out consequently, influence height-diameter ratio, diameter-thickness ratio and wind girder stiffness on top studied. Based results, regression equation for predicting maximum radial displacement proposed. Furthermore, accuracy proposed solution validated.

In this paper, the concept of double local lifetime control (DLLC) that enable lower forward voltage fast recovery diode is presented. Different from axial carrier and global used in conventional FRDs, DLLC FRDs have two independent life areas anode active region cathode respectively. The simulation results show diodes can effectively reduce drop reverse energy. Compared with diode, value reduced by 13.6%, energy 35.5%. And actual test same tendency a 13.5% reduction 21.2% compared to...