Results 21 to 30 of about 17,971 (165)
Data Preprocessing Method for Video Action Recognition Depth Models [PDF]
Video preprocessing operations, mainly including video frame sampling and data augmentation, are essential methods to improving the performance of deep video models in action recognition, and they have recently received increased attention. In this study,
Fengmin AN, Bingbing ZHANG, Wei DONG, Jianxin ZHANG
core +1 more source
Document-Level Event Detection Method Based on Information Aggregation and Data Augmentation [PDF]
Event detection is a key task in the field of natural language processing, aiming to identify event trigger words and correctly classify their event types.
TAN Lijun, HU Yanli, CAO Jianwei, TAN Zhen
core +1 more source
Self-supervised speech representation learning based on positive sample comparison and masking reconstruction [PDF]
, 2022 To solve the problem that existing contrastive prediction based self-supervised speech representation learning methods need to construct a large number of negative samples, and their performance depends on large training batches, requiring a lot of ...
Dan QU +4 more
core +1 more source
Image Data Augmentation Method for Random Channel Perturbation [PDF]
The simulation of object occlusion strategies in data augmentation sets all the pixels in the randomly cropped region of the input image to zero, which erases the effective texture features and leads to poor network generalization.
JIANG Wentao, LIU Yuwei, ZHANG Shengchong
core +1 more source
A Few‑Shot Learning Algorithm for Defect Image Generation and Data Augmentation Based on DID‑AugGAN [PDF]
To address the issues of low quality， lack of realism， and poor diversity in defect images generated by generative adversarial network （GAN） under small-sample conditions， this paper proposes a defect image generation algorithm， named defect image data ...
DENG Yafeng +3 more
core +7 more sources
Research on Fourier Augmented Unbiased Cross-Domain Object Detection [PDF]
The main purpose of unbiased cross-domain object detection is to utilize the knowledge of the source domain to the maximum extent through knowledge distillation, and reduce the cross-domain gap of the model through domain adaptation.
WANG Bing, XU Pei, ZHANG Xingpeng
core +1 more source
Few-shot cybersecurity event detection method by data augmentation with prompting question answering [PDF]
The cybersecurity field lacks sufficient annotated data for event recognition, and the scenarios and semantics are complex, making it difficult to construct accurate event recognition models.
BAI Qingchun +5 more
core +1 more source
Integrative Conservation, EarlyView.Urbanization may alter bird foraging. Austral Parakeets (Enicognathus ferrugineus) in Patagonia rely on introduced plants in urban areas, especially in winter, despite preferring natives in the wild. This seasonal reliance on introduced species highlights urbanization trade‐offs and underscores the need to manage green areas with native plants to ...
Rocío Bahía +2 more
wiley +1 more source
Gaoya dianqi, 2022 作为一种新兴的环保绝缘气体，七氟异丁腈（C4F7N）因其优异的绝缘性能和远低于六氟化硫（SF6）的全球增温潜势引起了研究人员的广泛关注。在C4F7N电子碰撞截面等基础数据不完整的情况下，文中基于气体流注放电判据，提出了C4F7N-CO2混合气体有效电离反应系数的计算方法，分析了C4F7N-CO2混合气体的临界击穿场强及绝缘性能的变化规律。研究结果表明，当C4F7N气体的热化系数为1.2或1.4时，C4F7N-CO2混合气体的临界击穿场强计算值与实验数据吻合较好。此外，C4F7N ...
邓云坤 +5 more
doaj
YOLO-MMCE algorithm for visual detection of engineering vehicles in high-consequence area [PDF]
Objective Engineering vehicles operating on large-scale construction sites in high-consequence areas pose severe safety hazards to buried pipelines. This paper addresses the shortcomings of current common techniques for detecting overlapped targets of ...
Degang WANG +3 more
core +1 more source