Paper Review and A. I. Contents
This section includes a review of some of the materials that I have presented since 2014 as graduate student and now as researcher in South Korea.
This material has been built on the basis of the referenced papers. I am uploading this content online for educational purposes to help students and researchers understand the concepts and fundamentals of some of the most relevant articles on deep learning and computer vision for object detection, image classification, instance-semantic segmentation, open-set recognition, reinforcement learning, activity recognition, generative adversarial networks, and others.
* Please use this material with the corresponding reference.
For any comments or concerns about the contents, feel free to email me at afuentes@jbnu.ac.kr.
More materials will be added soon...
2021
YOLO v4: Optimal Speed and Accuracy of Object Detection
Object detection
Universal Domain Adaptation
Open set domain adaptation
Large-Scale Object Detection in the Wild for Imbalanced Multi-Labels
Object detection
2020
Rethinking Pre-training and Self-training
Image classification
Multiple Object Forecasting: Predicting Future Object Locations in Diverse Environments
Object tracking
CSPNet: A new backbone that can enhance learning capabilities of CNN
Feature extractor
Attention is all you need
Spatio-temporal model (Fundament of Transformers)
Actor Conditioned Attention Maps for Video Action Detection
Video action detection
Fixing the train-test resolution discrepancy
FixEfficientNet
CBNet: A Novel Composite Backbone Network Architecture for Object Detection
CBNet (Object detection)
Class-Balanced Loss Based on Effective Number of Samples
Class imbalance
2019
EfficientDet: Scalable and Efficient Object Detection
EfficientDet (Object detection)
Bottom-up Object Detection by Grouping Extreme and Center Points
CornerNet
C2AE: Class Conditioned Auto-Encoder for Open-set Recognition
Open-set recognition
Cascade R-CNN: High Quality Object Detection and Instance Segmentation
Cascade R-CNN (Object detection and instance segmentation)
Libra R-CNN: Towards Balanced Learning for Object Detection
Object detection
Hierarchical Relational Networks for Group Activity Recognition and Retrieval
Activity recognition
Scale-Aware Trident Networks for Object Detection
TridentNet (Object detection)
SNIPER: Efficient Multi-Scale Training
SniperNet (Object detection)
Feature Selective Anchor-Free Module for Single-Shot Object Detection
Object detection
An Introduction to Deep Reinforcement Learning
Reinforcement Learning
Deformable Convolutional Networks
Optimization of feature extractors
2018
Pelee: A Real-Time Object Detection System on Mobile Devices
PeleeNet (Object detection)
A Review on Activity Recognition
Action recognition
MegDet: A large Mini-Batch Object Detector
Object detection
AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks
AttnGAN (Generative Adversarial Networks)
From Text-to-Image
Feature Generating Networks for Zero-Shot Learning
Zero-Shot Learning
MobileNetV2: Inverted Residuals and Linear Bottlenecks
Object detection
Taskonomy: Disentangling Task Transfer Learning
Transfer Learning
Analyzing Filters Toward Efficient ConvNet
Understanding ConvNets
Learning Transferable Architectures for Scalable Image Recognition
NASNet (Image Recognition)
FractalNet: Ultra-Deep Neural Networks Without Residuals
FractalNet (Image Recognition)
Single-Shot Refinement Neural Network for Object Detection
SSR (Object detection)
Focal Loss for Dense Object Detection
RetinaNet (Object detection)
2017
Understanding Deep Learning Requires Re-thinking Generalization
Generalization
Residual Attention Network for Image Classification
Image classification
Densely Connected Convolutional Networks
DenseNet
MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
Image classification and object detection
Detecting and Recognizing Human-Objects Interactions
Human-Object Interaction
YOLO9000: Better, Faster, Stronger
Object detection
Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
Fundament of GAN
Speed/accuracy trade-offs for modern convolutional object detectors
Tradeoff between accuracy and speed of deep nets for object detection
2016
SSD: Single Shot MultiBox Detector
SSD (Object detection)
ProNet: Learning to Propose Object-specific Boxes for Cascaded Neural Networks
Object detection
Learning Deep Features for Discriminative Localization
Object localization
Identity Mappings in Deep Residual Networks
Identity in residual networks
T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos
Object detection in videos
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Object detection
Pedestrian Detection aided by Deep Learning Semantic Tasks
Pedestrian detection
Recurrent Convolutional Neural Networks for Objects Recognition
RCNN (Object detection)
FlowNet: Learning Optical Flow with Convolutional Networks
Optical flow
2015
Traditional hand-crafted-based methods for pedestrian detection, motion estimation, optical flow.
Computing the Stereo Matching Cost with a Convolutional Neural Network
Stereo vision CNN
The Benefits of Dense Stereo for Pedestrian Detection
Dense stereo vision
Moving Objects Detection and Credal Boosting Based Recognition in Urban Environmets
Moving objects detection
Will the Pedestrian Cross? A study on Pedestrian Path Prediction
Pedestrian detection
Unsupervised flow-based motion analysis for an autonomous moving system
Motion analysis
Locating moving objects in car-driving sequences
Moving objects detection
Stixels motion estimation without optical flow computation
Motion estimation
Pedestrian detection from traffic scenes based on probabilistic models of the contour fragments
Pedestrian detection
Feature- and Depth-Supported Modified Total Variation Optical Flow for 3D Motion Field Estimation in Real Scenes
Optical flow
2014
Traditional hand-crafted-based methods for pedestrian detection, motion estimation, optical flow.
Moving Pedestrian Detection Based on Motion Segmentation
Motion segmentation
Motion Segmentation Using Optical Flow for Pedestrian Detection from Moving Vehicle
Pedestrian detection
Towards a Real-Time Pedestrian Detection based on a deformable template model
Pedestrian detection
New features and insights for pedestrian detection
Pedestrian detections
Pedestrian Detection: An Evaluation of the State of the Art
Pedestrian detection
Studying Relationships Between Human Gaze, Description, and Computer Vision
Human and computer vision