论文阅读杂记

论文阅读杂记

文章目录

      • 论文阅读杂记
        • 目标检测
        • Backbone
        • 细粒度识别
        • Attention机制
        • 剪枝
        • 超分辩率
        • 图像恢复
        • MIMO问题

目标检测

序号 名称 备注
1 learning Efficient Convolutional Networks through Network Slimming 基于BN的模型剪枝
2 FAIR1M: A Benchmark Dataset for Fine-grained Object Recognition in High-Resolution Remote Sensing Imagery FAIR1M:卫星遥感下的细粒度数据集
3 Rich feature hierarchies for accurate object detection and semantic segmentation Tech report(RCNN) RCNN
4 Fast RCNN Fast RCNN
5 Faster RCNN Faster RCNN
6 R2CNN: Rotational Region CNN for Orientation Robust Scene Text Detection R2CNN:text的斜框检测
7 Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition(SPP-net) SPP-net
8 ReDet:A Rotation-equivariant Detector for Aerial Object Detection ReDet
9 You Only Look Once: Unified, Real-Time Object Detection YOLO
10 YOLO9000: Better, Faster, Stronger YOLOv2
11 YOLOv3: An Incremental Improvement YOLOv3
12 YOLOv4: Optimal Speed and Accuracy of Object Detection YOLOv4
13 Cascade R-CNN: Delving into High Quality Object Detection Cascade RCNN
14 YOLOX: Exceeding YOLO Series in 2021 YOLOX
15 TPH-YOLOv5: Improved YOLOv5 Based on Transformer Prediction Head for Object Detection on Drone-captured Scenarios TPH-YOLOv5

Backbone

序号 名称 备注
1 Very Deep Convolutional Networks for Large-Scale Image Recognition VGG
2 OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks OverFeat
3 ResNet:Deep Residual Learning for Image Recognition ResNet
4 Aggregated Residual Transformations for Deep Neural Networks ResNeXt
5 MobileNets: Efficient Convolutional Neural Networks for Mobile Vision MobileNetv2
6 Searching for MobileNetV3 MobileNetv3
7 Densely Connected Convolutional Networks DenseNet
8 EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks EfficientNet
9 Swin Transformer简记-220112版 Swin Transformer
10 A ConvNet for the 2020s ConvNeXt

细粒度识别

序号 名称 备注
1 Mask-CNN: Localizing Parts and Selecting Descriptors for Fine-Grained Image Recognition Mask-CNN
2 FAIR1M: A Benchmark Dataset for Fine-grained Object Recognition in High-Resolution Remote Sensing Imagery FAIR1M:卫星遥感下的细粒度数据集
3 Selective Sparse Sampling for Fine-grained Image Recognition S3N,稀疏注意力
4 Filtration and Distillation: Enhancing Region Attention for Fine-Grained Visual Categorization FDL,RPN+知识蒸馏

Attention机制

序号 名称 备注
1 Residual Attention Network for Image Classification ResAN
2 CBAM: Convolutional Block Attention Module CBAM
3 BAM: Bottleneck Attention Module BAM
4 An Attention Module for Convolutional Neural Networks AW-Convlution

剪枝

序号 名称 备注
1 Learning both Weights and Connections for Efficient 直接用weight权重剪枝
2 learning Efficient Convolutional Networks through Network Slimming 基于BN的模型剪枝
3 Vision Transformer Pruning简记 FC的剪枝思想在Vision Transformer上的应用
4 Patch Slimming for Efficient Vision Transformers简记 自顶向下的Patch剪枝
5 Rethinking the Smaller-Norm-Less-Informative Assumption in Channel Pruning of Convolution Layers简记 也是BN的剪枝方法,类似Slimming
6 Coreset-Based Neural Network Compression简记 低秩分解
7 SELF-ADAPTIVE NETWORK PRUNING简记 注意力/自适应剪枝

超分辩率

序号 名称 备注
1 SRCNN简记 SRCNN
2 SRGAN简记 SRGAN
3 SCAN简记 SCAN
4 SwinIR简记 SwinIR

图像恢复

序号 名称 备注
1 MAE简记 kaiming

MIMO问题

序号 名称 备注
1 Convolutional Neural Network based Multiple-Rate Compressive Sensing for Massive MIMO CSI Feedback: Design, Simulation, and Analysis简记 19-CsiNet+
2 CLNet: Complex Input Lightweight Neural Network designed for Massive MIMO CSI Feedback简记 21-CLNet
3 Mino Channel Infomation Feedback Using Deep Recurrent Network简记 18,引入Lstm
4 Spatio-Temporal Representation With Deep Neural Recurrent Network in MIMO CSI Feedback简记 21-ConvlstmCsiNet

你可能感兴趣的:(论文阅读杂记,深度学习,1024程序员节,计算机视觉)