DenseNets have been shown to be a competitive model among recent convolutional network architectures. These networks utilize Dense Blocks, which are groups of densely connected layers where the output of a hidden layer is fed in as the input of every other layer following it. In this paper, we aim to improve certain aspects of DenseNet, especially when it comes to practicality. We introduce ParaNet, a new architecture that constructs three pipelines which allow for early inference. We additionally introduce a cascading mechanism such that different pipelines are able to share parameters, as well as logit matching between the outputs of the pipelines. We separately evaluate each of the newly introduced mechanisms of ParaNet, then evaluate our proposed architecture on CIFAR-100.
Dense Residual Network: Enhancing Global Dense Feature Flow for Character Recognition
Hefei University of Technology(合肥工业大学);Soochow University(苏州大学);Harbin Institute of Technology (Shenzhen)(哈尔滨工业大学(深圳));South China Normal University(华南师范大学);University of Science and Technology of China(中国科学技术大学)
Australian National University(澳大利亚国立大学);Data61, CSIRO(澳大利亚联邦科学与工业研究组织Data61);University of Oxford(牛津大学);EPFL(洛桑联邦理工学院);Alan Turing Institute(阿兰·图灵研究所)
AUNet: Attention-guided dense-upsampling networks for breast mass segmentation in whole mammograms
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences(中国科学院深圳先进技术研究院);Shandong University(山东大学);Guangdong General Hospital(广东省人民医院);Guangdong Academy of Medical Sciences(广东省医学科学院);Shanghai Jiao Tong University(上海交通大学);The University of Sydney(悉尼大学)
Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所);University of Chinese Academy of Sciences(中国科学院大学);University of Science and Technology of China(中国科学技术大学)
MSAD-Net: Multiscale and Spatial Attention-based Dense Network for Lung Cancer Classification
Computer Science and Engineering Department, NIIT University, Jaipur, India(尼特大学计算机科学与工程系,印度斋普尔);CSE department, Christ (Deemed to be University), Bangalore, India(基督大学(认定为大学)计算机科学与工程系,印度班加罗尔);Faculty of Computing, Engineering and the Built Environment, Birmingham City University, UK(伯明翰城市大学计算、工程与建筑环境学院,英国)
Memory Efficient Transformer Adapter for Dense Predictions
The Hong Kong University of Science and Technology(香港科技大学);InnoHK AI Chip Center for Smart Emerging Systems(InnoHK智能新兴系统AI芯片中心);Nanjing University(南京大学)
DenVisCoM: Dense Vision Correspondence Mamba for Efficient and Real-time Optical Flow and Stereo Estimation
Machine Intelligence Group, Birla Institute of Technology and Science, Pilani, Hyderabad Campus, India(机智人工智能组,比拉理工学院,比拉理工学院,海得拉巴校区,印度);Université Côte d’Azur, Inria, France(蔚蓝大学,法国国家信息与自动化研究所)
PAANet: Progressive Alternating Attention for Automatic Medical Image Segmentation
Indian Statistical Institute(印度统计研究所);Østfold University College(东福尔大学学院);SimulaMet;UiT The Arctic University of Norway(挪威北极大学);Oslo Metropolitan University(奥斯陆城市大学)
MTMamba++: Enhancing Multi-Task Dense Scene Understanding via Mamba-Based Decoders
Artificial Intelligence Thrust, The Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州)人工智能方向);Department of Computer Science and Engineering, The Hong Kong University of Science and Technology(香港科技大学计算机科学与工程系);SmartMore Corporation Limited(SmartMore有限公司)
MTMamba: Enhancing Multi-Task Dense Scene Understanding by Mamba-Based Decoders
The Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州));HKUST(GZ) - SmartMore Joint Lab(香港科技大学(广州)与视博医疗联合实验室);The Hong Kong University of Science and Technology(香港科技大学);Southern University of Science and Technology(南方科技大学);SmartMore(视博医疗)
Dynamic Block Matching to assess the longitudinal component of the dense motion field of the carotid artery wall in B-mode ultrasound sequences -- Association with coronary artery disease
Nara Institute of Science and Technology(奈良先端科学技术大学院大学);Technische Universität München(慕尼黑工业大学);National Cerebral and Cardiovascular Center(国立循环器病研究中心)
Limited View Tomographic Reconstruction Using a Deep Recurrent Framework with Residual Dense Spatial-Channel Attention Network and Sinogram Consistency
Yale University(耶鲁大学);Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所)
有限视图断层摄影重建旨在从有限数量的正弦图或投影视图重建断层摄影图像,这些正弦图或投影视图由减少辐射剂量或缩短扫描时间的稀疏视图或有限角度采集产生。然而,由于正弦图的不完整性,这种重建遭受高噪声和严重伪影。为了导出质量重建,先前的最先进的方法使用UNet类神经架构来从有限的视图数据直接预测全视图重建;但是这些方法在很大程度上保留了深度网络架构问题,并且不能保证重建图像的正弦图与采集的正弦图之间的一致性,从而导致非理想重建。在这项工作中,我们提出了一种新的经常性重建框架,堆叠相同的块多次。循环块由定制设计的剩余密集空间通道注意力网络组成。此外,我们开发了一个正弦图一致性层交织在我们的经常性框架,以确保采样正弦图是一致的经常性块的中间输出的正弦图。我们在两个数据集上评估我们的方法。我们在AAPM低剂量CT Grand Challenge数据集上的实验结果表明,我们的算法在有限角度重建(PSNR优于5dB)和稀疏视图重建(PSNR优于4dB)方面都比现有的最先进的神经方法实现了一致和显着的改进。此外,我们在深部病变数据集上的实验结果表明,我们的方法能够为8种主要病变类型生成高质量的重建。