Comput. Vis. Image Underst.最新文献_第3页

Transformer-based Image Generation from Scene Graphs 从场景图生成基于变压器的图像

Comput. Vis. Image Underst. Pub Date : 2023-03-08 DOI: 10.48550/arXiv.2303.04634

Renato Sortino, S. Palazzo, C. Spampinato

{"title":"Transformer-based Image Generation from Scene Graphs","authors":"Renato Sortino, S. Palazzo, C. Spampinato","doi":"10.48550/arXiv.2303.04634","DOIUrl":"https://doi.org/10.48550/arXiv.2303.04634","url":null,"abstract":"Graph-structured scene descriptions can be efficiently used in generative models to control the composition of the generated image. Previous approaches are based on the combination of graph convolutional networks and adversarial methods for layout prediction and image generation, respectively. In this work, we show how employing multi-head attention to encode the graph information, as well as using a transformer-based model in the latent space for image generation can improve the quality of the sampled data, without the need to employ adversarial models with the subsequent advantage in terms of training stability. The proposed approach, specifically, is entirely based on transformer architectures both for encoding scene graphs into intermediate object layouts and for decoding these layouts into images, passing through a lower dimensional space learned by a vector-quantized variational autoencoder. Our approach shows an improved image quality with respect to state-of-the-art methods as well as a higher degree of diversity among multiple generations from the same scene graph. We evaluate our approach on three public datasets: Visual Genome, COCO, and CLEVR. We achieve an Inception Score of 13.7 and 12.8, and an FID of 52.3 and 60.3, on COCO and Visual Genome, respectively. We perform ablation studies on our contributions to assess the impact of each component. Code is available at https://github.com/perceivelab/trf-sg2im","PeriodicalId":10549,"journal":{"name":"Comput. Vis. Image Underst.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80333562","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3

Low-light image enhancement by deep learning network for improved illumination map 基于深度学习网络的低照度图像增强

Comput. Vis. Image Underst. Pub Date : 2023-03-01 DOI: 10.2139/ssrn.4327727

Manli Wang, Jiayue Li, Changsen Zhang

引用次数: 0

Region selection for occluded person re-identification via policy gradient 基于策略梯度的闭塞人再识别区域选择

Comput. Vis. Image Underst. Pub Date : 2023-02-01 DOI: 10.2139/ssrn.4253486

Bolei Xu

引用次数: 2

Improved domain adaptive object detector via adversarial feature learning 基于对抗特征学习的改进域自适应目标检测器

Comput. Vis. Image Underst. Pub Date : 2023-02-01 DOI: 10.2139/ssrn.4013261

M. Marnissi, H. Fradi, A. Sahbani, N. Amara

引用次数: 0

Siamese Graph Attention Networks for robust visual object tracking 稳健视觉目标跟踪的Siamese图注意网络

Comput. Vis. Image Underst. Pub Date : 2023-02-01 DOI: 10.2139/ssrn.4067301

Junjie Lu, Shengyang Li, Weilong Guo, Manqi Zhao, Jian Yang, Yunfei Liu, Zhuang Zhou

引用次数: 0

Cross-domain multi-style merge for image captioning 图像字幕的跨域多样式合并

Comput. Vis. Image Underst. Pub Date : 2023-02-01 DOI: 10.2139/ssrn.4162675

Yiqun Duan, Zhen Wang, Li Yi, Jingya Wang

引用次数: 2

A unified RGB-T crowd counting learning framework 统一的RGB-T人群计数学习框架

Comput. Vis. Image Underst. Pub Date : 2023-01-01 DOI: 10.2139/ssrn.4098530

Siqi Gu, Z. Lian

引用次数: 2

Foreground discovery in streaming videos with dynamic construction of content graphs 动态构建内容图的流媒体视频前景发现

Comput. Vis. Image Underst. Pub Date : 2022-12-01 DOI: 10.2139/ssrn.4194725

Sepehr Farhand, G. Tsechpenakis

引用次数: 0

SSDA-YOLO: Semi-supervised Domain Adaptive YOLO for Cross-Domain Object Detection SSDA-YOLO:半监督域自适应YOLO跨域目标检测

Comput. Vis. Image Underst. Pub Date : 2022-11-04 DOI: 10.48550/arXiv.2211.02213

Huayi Zhou, Fei Jiang, Hongtao Lu

{"title":"SSDA-YOLO: Semi-supervised Domain Adaptive YOLO for Cross-Domain Object Detection","authors":"Huayi Zhou, Fei Jiang, Hongtao Lu","doi":"10.48550/arXiv.2211.02213","DOIUrl":"https://doi.org/10.48550/arXiv.2211.02213","url":null,"abstract":"Domain adaptive object detection (DAOD) aims to alleviate transfer performance degradation caused by the cross-domain discrepancy. However, most existing DAOD methods are dominated by outdated and computationally intensive two-stage Faster R-CNN, which is not the first choice for industrial applications. In this paper, we propose a novel semi-supervised domain adaptive YOLO (SSDA-YOLO) based method to improve cross-domain detection performance by integrating the compact one-stage stronger detector YOLOv5 with domain adaptation. Specifically, we adapt the knowledge distillation framework with the Mean Teacher model to assist the student model in obtaining instance-level features of the unlabeled target domain. We also utilize the scene style transfer to cross-generate pseudo images in different domains for remedying image-level differences. In addition, an intuitive consistency loss is proposed to further align cross-domain predictions. We evaluate SSDA-YOLO on public benchmarks including PascalVOC, Clipart1k, Cityscapes, and Foggy Cityscapes. Moreover, to verify its generalization, we conduct experiments on yawning detection datasets collected from various real classrooms. The results show considerable improvements of our method in these DAOD tasks, which reveals both the effectiveness of proposed adaptive modules and the urgency of applying more advanced detectors in DAOD. Our code is available on url{https://github.com/hnuzhy/SSDA-YOLO}.","PeriodicalId":10549,"journal":{"name":"Comput. Vis. Image Underst.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84102224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 10

Feature reconstruction and metric based network for few-shot object detection 基于特征重构和度量的小目标检测网络

Comput. Vis. Image Underst. Pub Date : 2022-11-01 DOI: 10.2139/ssrn.4013260

Yuewen Li, W. Feng, Shuchang Lyu, Q. Zhao

引用次数: 2