Msmsfnet: a multi-stream and multi-scale fusion net for edge detection
School of Architecture and Urban Planning, Shenzhen University, Shenzhen, China(深圳大学建筑与城市规划学院);Ministry of Natural Resources (MNR) Key Laboratory for Geo-Environmental Monitoring of Great Bay Area, Shenzhen, China(自然资源部粤港澳大湾区生态环境监测重点实验室);Guangdong Key Laboratory of Urban Informatics, Shenzhen, China(广东省城市信息学重点实验室);Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ), Shenzhen, China(广东省人工智能与数字经济实验室(深圳));China Mobile Internet Company Limited, Guangzhou, China(中国移动互联网有限公司);Key Laboratory for Information Science of Electromagnetic Waves (Ministry of Education), School of Information Science and Technology, Fudan University, Shanghai, China(信息电磁波科学国家重点实验室(教育部));School of Remote Sensing and Information Engineering, Wuhan University, Wuhan, China(武汉大学遥感与信息工程学院)
Chenguang Liu, Chisheng Wang, Feifei Dong, Xiayang Xiao, Xin Su, Chuanhua Zhu, Dejin Zhang, Qingquan Li
针对边缘检测依赖 ImageNet 预训练权重的问题,本文提出多流多尺度融合网络 msmsfnet,并证明其从零训练时在三个公开数据集及 SAR 图像上优于现有方法。
摘要:
边缘检测是计算机视觉中的一个长期存在的问题。尽管现有算法具有较高效率,但其性能在很大程度上依赖于骨干网络在 ImageNet 数据集上的预训练权重。以往方法对预训练权重的使用显著增加了在不依赖现有训练良好的 ImageNet 模型的情况下设计新的边缘检测模型的难度,因为在 ImageNet 数据集上预训练模型成本高昂,而且为了保证比较的公平性,这种预训练又成为必需。此外,预训练加微调策略并不总是有效,有时甚至无法使用。例如,由于光学图像与合成孔径雷达(SAR)图像在统计特性上存在显著差异,ImageNet 数据集上的预训练权重不太可能有助于 SAR 图像中的边缘检测。而且,在 SAR 图像处理领域,没有任何数据集具有可与 ImageNet 数据集相比的规模。在本研究中,我们研究了当前最先进的基于深度学习的边缘检测器在公开可用数据集上从零开始训练时能够达到的性能,并设计了一种新的网络架构,即用于边缘检测的多流多尺度融合网络(msmsfnet)。我们在实验中表明,通过从零开始训练所有模型,我们的模型在三个公开可用数据集上优于最先进的边缘检测器。我们还证明了模型在没有可用预训练权重的 SAR 图像边缘检测中的效率。最后,我们表明,在使用预训练权重时,我们的模型能够在 BSDS500 数据集上取得具有竞争力的性能。
英文摘要:
Edge detection is a long-standing problem in computer vision. Despite the efficiency of existing algorithms, their performance, however, rely heavily on the pre-trained weights of the backbone network on the ImageNet dataset. The use of pre-trained weights in previous methods significantly increases the difficulty to design new models for edge detection without relying on existing well-trained ImageNet models, as pre-training the model on the ImageNet dataset is expensive and becomes compulsory to ensure the fairness of comparison. Besides, the pre-training and fine-tuning strategy is not always useful and sometimes even inaccessible. For instance, the pre-trained weights on the ImageNet dataset are unlikely to be helpful for edge detection in Synthetic Aperture Radar (SAR) images due to strong differences in the statistics between optical images and SAR images. Moreover, no dataset has comparable size to the ImageNet dataset for SAR image processing. In this work, we study the performance achievable by state-of-the-art deep learning based edge detectors in publicly available datasets when they are trained from scratch, and devise a new network architecture, the multi-stream and multi-scale fusion net (msmsfnet), for edge detection. We show in our experiments that by training all models from scratch, our model outperforms state-of-the-art edge detectors in three publicly available datasets. We also demonstrate the efficiency of our model for edge detection in SAR images, where no useful pre-trained weight is available. Finally, We show that our model is able to achieve competitive performance on the BSDS500 dataset when the pre-trained weights are used.
TB-Net: A Three-Stream Boundary-Aware Network for Fine-Grained Pavement Disease Segmentation
Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所);School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院)
Two-stream Encoder-Decoder Network for Localizing Image Forgeries
Indian Institute of Technology Guwahati(印度理工学院古瓦哈蒂分校);Department of Electronics and Electrical Engineering(电子与电气工程系);Department of Electronics and Electrical Engineering, Indian Institute of Technology Guwahati(电子与电气工程系,印度理工学院古瓦哈蒂分校);Indian Institute of Technology Guwahati, Assam-781039, India(印度理工学院古瓦哈蒂分校,印度阿萨姆邦-781039)
Two-Stream Multi-Task Network for Fashion Recognition
Beihang University(北京航空航天大学);Inception Institute of Artificial Intelligence(Inception人工智能研究院);Guangdong University of Petrochemical Technology(广东石油化工学院)