zhaixueming / cv_paperdaily

CV 每日论文阅读笔记（日更）

Notice！

Keep update. On vacation... Busy withKeep up to date recent things , I'll update without day.

CV--PaperDaily

创建人	知乎论文阅读专栏	个人博客	其他相关链接
ming71	论文笔记入口	chaser	CSDN

Update CV papers here everday .
The content includes but is not limited to Object detection , Semantic segmentation , and other papers about deep learning . Most of papers are published in recent two years
Your comments are welcome , and you can e-mail me by [email protected] .

Paper reading divided by conference

AAAI

M2Det: A Single-Shot Object Detector based on Multi-Level Feature Pyramid

---

CVPR

• Assisted Excitation of Activations: A Learning Technique to Improve Object
• Borrow from Anywhere Pseudo Multi-modal Object Detection in Thermal Imagery
• Cascade R-CNN: Delving into High Quality Object Detection
• Feature Pyramid Networks for Object Detection
• Not All Pixels Are Equal: Difficulty-Aware Semantic Segmentation via Deep Layer Cascade
• Path Aggregation Network for Instance Segmentation
• Region Proposal by Guided Anchoring
• Scale-Transferrable Object Detection
• DOTA: A Large-scale Dataset for Object Detection in Aerial Images
• R2CNN: Rotational Region CNN for Orientation Robust Scene Text Detection

---

ECCV

• DetNet: A Backbone network for Object Detection
• Receptive Field Block Net for Accurate and Fast Object Detection

---

ICCV

• Focal Loss for Dense Object Detection

---

ICML

• Making Convolutional Networks Shift-Invariant Again

---

ICLR

• Why do deep convolutional networks generalize so poorly to small image transformations?
• Dataset Augmentationin In Feature Space
• ImageNet-trained CNNs are biased towards texture: increasing shape bias improves accuracy and robustness

---

ACCV

• Reverse Densely Connected Feature Pyramid Network for Object Detection

---

IJCV

• What Makes Good Synthetic Training Data for Learning Disparity and Optical Flow Estimation?

---

ArXiv

• FSSD: Feature Fusion Single Shot Multibox Detector
• MDSSD: Multi-scale Deconvolutional Single Shot Detector for Small Objects
• Scale-Aware Trident Networks for Object Detection
• ThunderNet:Towards Real-time Generic Object Detection
• MMDetection: Open MMLab Detection Toolbox and Benchmark
• Double-Head RCNN: Rethinking Classification and Localization for Object Detection
• Learning Data Augmentation Strategies for Object Detection
• A Preliminary Study on Data Augmentation of Deep Learning for Image Classification
• Improved Regularization of Convolutional Neural Networks with Cutout
• Data Augmentation by Pairing Samples for Images Classification
• How much real data do we actually need: Analyzing object detection performance using synthetic and real data
• Bag of Freebies for Training Object Detection Neural Networks
• The Effectiveness of Data Augmentation in Image Classification using Deep Learning
• Natural Adversarial Examples
• Recent Advances in Deep Learning for Object Detection

---

Ohters

• (IEEE Acess) Smart Augmentation: Learning an Optimal Data Augmentation Strategy
• (ICANN) Further advantages of data augmentation on convolutional neural networks
• (WACV) Understanding Convolution for Semantic Segmentation
• (BMCV)Enhancement of SSD by concatenating feature maps for object detection
• (Journal of Big Data) A survey on Image Data Augmentation for Deep Learning
• (DICTA) Understanding data augmentation for classification: when to warp?

---

笔记效果

Written by markdown .Displayed as follow :