This repository contains my paper reading notes on deep learning and machine learning. It is inspired by Denny Britz and Daniel Takeshi.

New year resolution for 2020: read at least three paper a week and a high a high quality github repo a month!

The summary of the papers read in 2019 can be found here on Towards Data Science.

The sections below records paper reading activity in chronological order. See notes organized according to subfields here (up to 06-2019).

Here is a list of trustworthy sources of papers in case I ran out of papers to read.

The list of resource in this link talks about various topics in Autonomous Driving.

• simple-faster-rcnn-pytorch (2.1k star) [Notes]
• YOLACT/YOLACT++ (2.1k star)
• MonoLoco (131 star)
• A Baseline for 3D Multi-Object Tracking (548 star)
• ROLO: recurrent YOLO
• point rend
• Carla data export
• openpilot
• 3D Lane Dataset

• Modern C++ for computer vision
• SLAM by Cyrill Stachniss

• ChauffeurNet: Learning to Drive by Imitating the Best and Synthesizing the Worst [Notes] RSS 2019 [Waymo]
• IntentNet: Learning to Predict Intention from Raw Sensor Data [Notes] CoRL 2018 [Uber ATG, perception and prediction, Lidar+Map]
• RoR: Rules of the Road: Predicting Driving Behavior with a Convolutional Model of Semantic Interactions [Notes] CVPR 2019 [Zoox]
• MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction [Notes] CoRL 2019 [Waymo, authors from RoR and ChauffeurNet]
• NMP: End-to-end Interpretable Neural Motion Planner [Notes] CVPR 2019 oral [Uber ATG]
• Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks [Notes] ICRA 2019 [Multimodal]
• Jointly Learnable Behavior and Trajectory Planning for Self-Driving Vehicles IROS 2019 Oral [Uber ATG, behavioral planning, motion planning]
• Baidu Apollo EM Motion Planner
• D3VO: Deep Depth, Deep Pose and Deep Uncertainty for Monocular Visual Odometry CVPR 2020 [Daniel Cremer, TUM]
• Visual Odometry Revisited: What Should Be Learnt? ICRA 2020
• Designing Network Design Spaces CVPR 2020
• Moco2: Improved Baselines with Momentum Contrastive Learning
• ONCE: Incremental Few-Shot Object Detection CVPR 2020
• TensorMask: A Foundation for Dense Object Segmentation [Notes] ICCV 2019 [single-stage instance seg]
• PolarMask: Single Shot Instance Segmentation with Polar Representation [Notes] CVPR 2020 oral [single-stage instance seg]
• SOLO: Segmenting Objects by Locations [Notes] [single-stage instance seg]
• SOLOv2: Dynamic, Faster and Stronger [Notes] [single-stage instance seg]
• Conditional Convolutions for Instance Segmentation
• EfficientDet: Scalable and Efficient Object Detection
• RetinaTrack: Online Single Stage Joint Detection and Tracking CVPR 2020
• Reinforced Feature Points: Optimizing Feature Detection and Description for a High-Level Task CVPR 2020 oral [Eric Brachmann, ngransac]
• Detecting Lane and Road Markings at A Distance with Perspective Transformer Layers (3D LLD)
• Gliding vertex on the horizontal bounding box for multi-oriented object detection
• PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization [Notes] ICCV 2015
• PoseNet2: Modelling Uncertainty in Deep Learning for Camera Relocalization ICRA 2016
• PoseNet3: Geometric Loss Functions for Camera Pose Regression with Deep Learning CVPR 2017
• Learning Depth-Guided Convolutions for Monocular 3D Object Detection (Mono3D)
• Unsupervised Feature Learning via Non-Parametric Instance-level Discrimination CVPR 2018 spotlight (Stella Yu)
• Render for CNN: Viewpoint Estimation in Images Using CNNs Trained with Rendered 3D Model Views ICCV 2015 Oral
• Towards High Performance Video Object Detection [Notes] CVPR 2018
• Towards High Performance Video Object Detection for Mobiles [Notes]
• Theoretical insights into the optimization landscape of over-parameterized shallow neural networks TIP 2018
• The Power of Interpolation: Understanding the Effectiveness of SGD in Modern Over-parametrized Learning ICML 2018
• SGD on Neural Networks Learns Functions of Increasing Complexity NIPS 2019 (SGD learns a linear classifier first)
• Pay attention to the activations: a modular attention mechanism for fine-grained image recognition
• Computer Vision for Autonomous Vehicles: Problems, Datasets and State of the Art (latest update in Dec 2019)
• Simultaneous Identification and Tracking of Multiple People Using Video and IMUs CVPR 2019
• Detect-and-Track: Efficient Pose Estimation in Videos
• TrackNet: Simultaneous Object Detection and Tracking and Its Application in Traffic Video Analysis
• Video Action Transformer Network CVPR 2019 oral
• Online Real-time Multiple Spatiotemporal Action Localisation and Prediction ICCV 2017
• 多目标跟踪 近年论文及开源代码汇总
• Long-Term On-Board Prediction of People in Traffic Scenes under Uncertainty CVPR 2018 [on-board bbox prediction]
• Egocentric Vision-based Future Vehicle Localization for Intelligent Driving Assistance Systems (Honda)
• Unsupervised Traffic Accident Detection in First-Person Videos IROS 2019 (Honda)
• NEMO: Future Object Localization Using Noisy Ego Priors (Honda)
• Robust Aleatoric Modeling for Future Vehicle Localization (perspective)
• Multiple Object Forecasting: Predicting Future Object Locations in Diverse Environments WACV 2020 (perspective bbox, pedestrian)
• TSM: Temporal Shift Module for Efficient Video Understanding ICCV 2019 (Song Han)
• AGSS-VOS: Attention Guided Single-Shot Video Object Segmentation ICCV 2019
• One-Shot Video Object Segmentation CVPR 2017
• Looking Fast and Slow: Memory-Guided Mobile Video Object Detection CVPR 2018
• Real-Time Seamless Single Shot 6D Object Pose Prediction CVPR 2018
• Learn to Combine Modalities in Multimodal Deep Learning (sensor fusion, general DL)
• PointRNN: Point Recurrent Neural Network for Moving Point Cloud Processing
• Using panoramic videos for multi-person localization and tracking in a 3D panoramic coordinate
• UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction [Notes] (dimension reduction, better than t-SNE)
• Deep Cuboid Detection: Beyond 2D Bounding Boxes (Magic Leap)
• Viewpoints and Keypoints (Malik)
• Lifting Object Detection Datasets into 3D (PASCAL)
• 3D Object Class Detection in the Wild (keypoint based)
• Fast Single Shot Detection and Pose Estimation 3DV 2016 (SSD + pose, Wei Liu)
• Towards Scene Understanding with Detailed 3D Object Representations IJCV 2014 (keypoint, 3D bbox annotation)
• Virtual KITTI 2
• Deep Supervision with Shape Concepts for Occlusion-Aware 3D Object Parsing CVPR 2017
• A Mixed Classification-Regression Framework for 3D Pose Estimation from 2D Images BMVC 2018 (multi-bin, what's new?)
• Safe Trajectory Generation For Complex Urban Environments Using Spatio-temporal Semantic Corridor LRA 2019 [Motion planning]
• Efficient Uncertainty-aware Decision-making for Automated Driving Using Guided Branching ICRA 2020 [Motion planning]
• DAgger: Driving Policy Transfer via Modularity and Abstraction CoRL 2018 [DAgger, Immitation Learning]
• Exploiting Sparse Semantic HD Maps for Self-Driving Vehicle Localization IROS 2019 oral [Uber ATG, metadata, localization]

• VPGNet: Vanishing Point Guided Network for Lane and Road Marking Detection and Recognition [Notes] ICCV 2017
• Which Tasks Should Be Learned Together in Multi-task Learning? [Notes] [Stanford]
• Multi-Task Learning as Multi-Objective Optimization NeurIPS 2018
• Taskonomy: Disentangling Task Transfer Learning [Notes] CVPR 2018
• Rethinking ImageNet Pre-training [Notes] ICCV 2019 [Kaiming He]
• Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network CVPR 2016 [channel-to-pixel]
• UnsuperPoint: End-to-end Unsupervised Interest Point Detector and Descriptor [Notes] [superpoint]
• KP2D: Neural Outlier Rejection for Self-Supervised Keypoint Learning [Notes] ICLR 2020 (pointNet)
• KP3D: Self-Supervised 3D Keypoint Learning for Ego-motion Estimation [Notes] [Toyota, superpoint]
• NG-RANSAC: Neural-Guided RANSAC: Learning Where to Sample Model Hypotheses [Notes] ICCV 2019 [pointNet]
• Learning to Find Good Correspondences [Notes] CVPR 2018 Oral (pointNet)
• RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving [Notes] [Huawei, Mono3D]
• DSP: Monocular 3D Object Detection with Decoupled Structured Polygon Estimation and Height-Guided Depth Estimation [Notes] AAAI 2020 (SenseTime, Mono3D)
• Robust Lane Detection from Continuous Driving Scenes Using Deep Neural Networks (LLD, LSTM)
• LaneNet: Towards End-to-End Lane Detection: an Instance Segmentation Approach [Notes] IV 2018 (LaneNet)
• 3D-LaneNet: End-to-End 3D Multiple Lane Detection [Notes] ICCV 2019
• Semi-Local 3D Lane Detection and Uncertainty Estimation [Notes] [GM Israel, 3D LLD]
• Gen-LaneNet: A Generalized and Scalable Approach for 3D Lane Detection [Notes] [Apollo, 3D LLD]
• Long-Term On-Board Prediction of People in Traffic Scenes under Uncertainty CVPR 2018 [Egocentric prediction]

• Associative Embedding: End-to-End Learning for Joint Detection and Grouping [Notes] NIPS 2017
• Pixels to Graphs by Associative Embedding [Notes] NIPS 2017
• Social LSTM: Human Trajectory Prediction in Crowded Spaces [Notes] CVPR 2017
• Online Video Object Detection using Association LSTM [Notes] [single stage, recurrent]
• SuperPoint: Self-Supervised Interest Point Detection and Description [Notes] CVPR 2018 (channel-to-pixel, deep SLAM, Magic Leap)
• PointRend: Image Segmentation as Rendering [Notes] CVPR 2020 Oral [Kaiming He, FAIR]
• Multigrid: A Multigrid Method for Efficiently Training Video Models [Notes] CVPR 2020 Oral [Kaiming He, FAIR]
• GhostNet: More Features from Cheap Operations [Notes] CVPR 2020
• FixRes: Fixing the train-test resolution discrepancy [Notes] NIPS 2019 [FAIR]
• VirtualCam: Single-Stage Monocular 3D Object Detection with Virtual Cameras [Notes] [Mapillary, Mono3D]
• Amodal Completion and Size Constancy in Natural Scenes [Notes] ICCV 2015 (Amodal completion)
• MoCo: Momentum Contrast for Unsupervised Visual Representation Learning [Notes] CVPR 2020 Oral [FAIR, Kaiming He]

• Double Descent: Reconciling modern machine learning practice and the bias-variance trade-of [Notes] PNAS 2019
• Deep Double Descent: Where Bigger Models and More Data Hurt [Notes]
• Visualizing the Loss Landscape of Neural Nets NIPS 2018
• The ApolloScape Open Dataset for Autonomous Driving and its Application CVPR 2018 (dataset)
• ApolloCar3D: A Large 3D Car Instance Understanding Benchmark for Autonomous Driving [Notes] CVPR 2019
• Part-level Car Parsing and Reconstruction from a Single Street View [Notes] [Baidu]
• 6D-VNet: End-to-end 6DoF Vehicle Pose Estimation from Monocular RGB Images [Notes] CVPR 2019
• RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving [Notes]
• DORN: Deep Ordinal Regression Network for Monocular Depth Estimation [Notes] CVPR 2018
• D&T: Detect to Track and Track to Detect [Notes] ICCV 2017 (from Feichtenhofer)
• CRF-Net: A Deep Learning-based Radar and Camera Sensor Fusion Architecture for Object Detection [Notes] SDF 2019 (radar detection)
• RVNet: Deep Sensor Fusion of Monocular Camera and Radar for Image-based Obstacle Detection in Challenging Environments [Notes] PSIVT 2019
• RRPN: Radar Region Proposal Network for Object Detection in Autonomous Vehicles [Notes] ICIP 2019
• ROLO: Spatially Supervised Recurrent Convolutional Neural Networks for Visual Object Tracking [Notes] ISCAS 2016
• Recurrent SSD: Recurrent Multi-frame Single Shot Detector for Video Object Detection [Notes] BMVC 2018 (Mitsubishi)
• Recurrent RetinaNet: A Video Object Detection Model Based on Focal Loss [Notes] ICONIP 2018 (single stage, recurrent)
• Actions as Moving Points [Notes] [not suitable for online]
• The PREVENTION dataset: a novel benchmark for PREdiction of VEhicles iNTentIONs [Notes] ITSC 2019
• Semi-Automatic High-Accuracy Labelling Tool for Multi-Modal Long-Range Sensor Dataset [Notes] IV 2018
• Astyx dataset: Automotive Radar Dataset for Deep Learning Based 3D Object Detection [Notes] EuRAD 2019 (Astyx)
• Astyx camera radar: Deep Learning Based 3D Object Detection for Automotive Radar and Camera [Notes] EuRAD 2019 (Astyx)

• How Do Neural Networks See Depth in Single Images? [Notes] ICCV 2019
• Self-supervised Sparse-to-Dense: Self-supervised Depth Completion from LiDAR and Monocular Camera ICRA 2019 (depth completion)
• DC: Depth Coefficients for Depth Completion [Notes] CVPR 2019 [Xiaoming Liu, Multimodal]
• Parse Geometry from a Line: Monocular Depth Estimation with Partial Laser Observation [Notes] ICRA 2017
• PointPainting: Sequential Fusion for 3D Object Detection (nuscenece)
• VO-Monodepth: Enhancing self-supervised monocular depth estimation with traditional visual odometry [Notes] 3DV 2019 (sparse to dense)
• Probabilistic Object Detection: Definition and Evaluation [Notes]
• The Fishyscapes Benchmark: Measuring Blind Spots in Semantic Segmentation [Notes] ICCV 2019
• On Calibration of Modern Neural Networks [Notes] ICML 2017 (Weinberger)
• Extreme clicking for efficient object annotation [Notes] ICCV 2017
• Radar and Camera Early Fusion for Vehicle Detection in Advanced Driver Assistance Systems [Notes] NeurIPS 2019 (radar)
• Deep Active Learning for Efficient Training of a LiDAR 3D Object Detector [Notes] IV 2019
• C3DPO: Canonical 3D Pose Networks for Non-Rigid Structure From Motion [Notes] ICCV 2019
• YOLACT: Real-time Instance Segmentation [Notes] ICCV 2019 [single-stage instance seg]
• YOLACT++: Better Real-time Instance Segmentation [single-stage instance seg]

• Review of Image and Feature Descriptors
• Vehicle Detection With Automotive Radar Using Deep Learning on Range-Azimuth-Doppler Tensors [Notes] ICCV 2019
• GPP: Ground Plane Polling for 6DoF Pose Estimation of Objects on the Road [Notes] IV 2020 [UCSD, Trevidi, mono 3DOD]
• MVRA: Multi-View Reprojection Architecture for Orientation Estimation [Notes] ICCV 2019
• YOLOv3: An Incremental Improvement
• Gaussian YOLOv3: An Accurate and Fast Object Detector Using Localization Uncertainty for Autonomous Driving [Notes] ICCV 2019 (Detection with Uncertainty)
• Bayesian YOLOv3: Uncertainty Estimation in One-Stage Object Detection [Notes] [DriveU]
• Towards Safe Autonomous Driving: Capture Uncertainty in the Deep Neural Network For Lidar 3D Vehicle Detection [Notes] ITSC 2018 (DriveU)
• Leveraging Heteroscedastic Aleatoric Uncertainties for Robust Real-Time LiDAR 3D Object Detection [Notes] IV 2019 (DriveU)
• Can We Trust You? On Calibration of a Probabilistic Object Detector for Autonomous Driving [Notes] IROS 2019 (DriveU)
• LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving [Notes] CVPR 2019 (uncertainty)
• LaserNet KL: Learning an Uncertainty-Aware Object Detector for Autonomous Driving [Notes] [LaserNet with KL divergence]
• IoUNet: Acquisition of Localization Confidence for Accurate Object Detection [Notes] ECCV 2018
• gIoU: Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression [Notes] CVPR 2019
• KL Loss: Bounding Box Regression with Uncertainty for Accurate Object Detection [Notes] CVPR 2019
• CAM-Convs: Camera-Aware Multi-Scale Convolutions for Single-View Depth [Notes] CVPR 2019
• BayesOD: A Bayesian Approach for Uncertainty Estimation in Deep Object Detectors [Notes]
• TW-SMNet: Deep Multitask Learning of Tele-Wide Stereo Matching [Notes] ICIP 2019
• Accurate Uncertainties for Deep Learning Using Calibrated Regression [Notes] ICML 2018
• Calibrating Uncertainties in Object Localization Task [Notes] NIPS 2018
• SMWA: On the Over-Smoothing Problem of CNN Based Disparity Estimation [Notes] ICCV 2019 [Multimodal, depth estimation]
• Sparse-to-Dense: Depth Prediction from Sparse Depth Samples and a Single Image [Notes] ICRA 2018 (depth completion)

• Review of monocular object detection
• Review of 2D 3D contraints in Mono 3DOD
• MonoGRNet 2: Monocular 3D Object Detection via Geometric Reasoning on Keypoints [Notes] [estimates depth from keypoints]
• Deep MANTA: A Coarse-to-fine Many-Task Network for joint 2D and 3D vehicle analysis from monocular image [Notes] CVPR 2017
• SS3D: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss [Notes] [rergess distance from images, centernet like]
• GS3D: An Efficient 3D Object Detection Framework for Autonomous Driving [Notes] CVPR 2019
• M3D-RPN: Monocular 3D Region Proposal Network for Object Detection [Notes] ICCV 2019 (Xiaoming Liu)
• TLNet: Triangulation Learning Network: from Monocular to Stereo 3D Object Detection [Notes] CVPR 2019
• A Survey on 3D Object Detection Methods for Autonomous Driving Applications [Notes] TITS 2019 [Review]
• BEV-IPM: Deep Learning based Vehicle Position and Orientation Estimation via Inverse Perspective Mapping Image [Notes] IV 2019
• ForeSeE: Task-Aware Monocular Depth Estimation for 3D Object Detection [Notes] AAAI 2020 oral [successor to pseudo-lidar, mono 3DOD SOTA]
• Obj-dist: Learning Object-specific Distance from a Monocular Image [Notes] ICCV 2019 (xmotors.ai + NYU)
• DisNet: A novel method for distance estimation from monocular camera [Notes] IROS 2018
• BirdGAN: Learning 2D to 3D Lifting for Object Detection in 3D for Autonomous Vehicles [Notes] IROS 2019
• Shift R-CNN: Deep Monocular 3D Object Detection with Closed-Form Geometric Constraints [Notes] ICIP 2019
• 3D-RCNN: Instance-level 3D Object Reconstruction via Render-and-Compare [Notes] CVPR 2018
• Deep Optics for Monocular Depth Estimation and 3D Object Detection [Notes] ICCV 2019
• MonoLoco: Monocular 3D Pedestrian Localization and Uncertainty Estimation [Notes] ICCV 2019
• Joint Monocular 3D Vehicle Detection and Tracking [Notes] ICCV 2019 (Berkeley DeepDrive)
• CasGeo: 3D Bounding Box Estimation for Autonomous Vehicles by Cascaded Geometric Constraints and Depurated 2D Detections Using 3D Results [Notes]

• Slimmable Neural Networks [Notes] ICLR 2019
• Universally Slimmable Networks and Improved Training Techniques [Notes] ICCV 2019
• AutoSlim: Towards One-Shot Architecture Search for Channel Numbers
• Once for All: Train One Network and Specialize it for Efficient Deployment
• DOTA: A Large-scale Dataset for Object Detection in Aerial Images [Notes] CVPR 2018 (rotated bbox)
• RoiTransformer: Learning RoI Transformer for Oriented Object Detection in Aerial Images [Notes] CVPR 2019 (rotated bbox)
• RRPN: Arbitrary-Oriented Scene Text Detection via Rotation Proposals TMM 2018
• R2CNN: Rotational Region CNN for Orientation Robust Scene Text Detection (rotated bbox)
• TI white paper: Webinar: mmWave Radar for Automotive and Industrial applications [Notes] [TI, radar]
• Federated Learning: Strategies for Improving Communication Efficiency [Notes] NIPS 2016
• sort: Simple Online and Realtime Tracking [Notes] ICIP 2016
• deep-sort: Simple Online and Realtime Tracking with a Deep Association Metric [Notes]
• MT-CNN: Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks [Notes] SPL 2016 (real time, facial landmark)
• RetinaFace: Single-stage Dense Face Localisation in the Wild [Notes] [joint object and landmark detection]
• SC-SfM-Learner: Unsupervised Scale-consistent Depth and Ego-motion Learning from Monocular Video [Notes] NIPS 2019
• SiamMask: Fast Online Object Tracking and Segmentation: A Unifying Approach CVPR 2019 (tracking, segmentation, label propagation)
• Review of Kálmán Filter (from Tim Babb, Pixar Animation) [Notes]
• R-FCN: Object Detection via Region-based Fully Convolutional Networks [Notes] NIPS 2016
• Guided backprop: Striving for Simplicity: The All Convolutional Net [Notes] ICLR 2015
• Occlusion-Net: 2D/3D Occluded Keypoint Localization Using Graph Networks [Notes] CVPR 2019
• Boxy Vehicle Detection in Large Images [Notes] ICCV 2019
• FQNet: Deep Fitting Degree Scoring Network for Monocular 3D Object Detection [Notes] CVPR 2019 (Mono 3DOD, Jiwen Lu)

• Mono3D: Monocular 3D Object Detection for Autonomous Driving [Notes] CVPR2016
• MonoDIS: Disentangling Monocular 3D Object Detection [Notes] ICCV 2019
• Pseudo lidar-e2e: Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud [Notes] ICCV 2019 (pseudo-lidar with 2d and 3d consistency loss, better than PL and worse than PL++, SOTA for pure mono3D)
• MonoGRNet: A Geometric Reasoning Network for Monocular 3D Object Localization [Notes] AAAI 2019 (SOTA of Mono3DOD, MLF < MonoGRNet < Pseudo-lidar)
• MLF: Multi-Level Fusion based 3D Object Detection from Monocular Images [Notes] CVPR 2018 (precursor to pseudo-lidar)
• ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape [Notes] CVPR 2019
• Accurate Monocular 3D Object Detection via Color-Embedded 3D Reconstruction for Autonomous Driving [Notes] ICCV 2019 [similar to pseudo-lidar, color-enhanced]
• Mono3D++: Monocular 3D Vehicle Detection with Two-Scale 3D Hypotheses and Task Priors [Notes] (from Stefano Soatto) AAAI 2019
• Deep Metadata Fusion for Traffic Light to Lane Assignment [Notes] IEEE RA-L 2019 (traffic lights association)
• Automatic Traffic Light to Ego Vehicle Lane Association at Complex Intersections ITSC 2019 (traffic lights association)
• Distant Vehicle Detection Using Radar and Vision[Notes] ICRA 2019 [radar, vision, radar tracklets fusion]
• Distance Estimation of Monocular Based on Vehicle Pose Information [Notes]
• Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics [Notes] CVPR 2018 (Alex Kendall)
• GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks [Notes] ICML 2018 (multitask)
• DTP: Dynamic Task Prioritization for Multitask Learning [Notes] ECCV 2018 [multitask, Stanford]
• Will this car change the lane? - Turn signal recognition in the frequency domain [Notes] IV 2014
• Complex-YOLO: Real-time 3D Object Detection on Point Clouds [Notes] (BEV detection only)
• Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds CVPR 2019 (sensor fusion and tracking)
• An intriguing failing of convolutional neural networks and the CoordConv solution [Notes] NIPS 2018

• Deep Parametric Continuous Convolutional Neural Networks [Notes] CVPR 2018 (@Uber, sensor fusion)
• ContFuse: Deep Continuous Fusion for Multi-Sensor 3D Object Detection [Notes] ECCV 2018 [Uber ATG, sensor fusion, BEV]
• Fast and Furious: Real Time End-to-End 3D Detection, Tracking and Motion Forecasting with a Single Convolutional Net [Notes] CVPR 2018 oral [lidar only, perception and prediction]
• Depth from Videos in the Wild: Unsupervised Monocular Depth Learning from Unknown Cameras [Notes] ICCV 2019 [monocular depth estimation, intrinsic estimation, SOTA]
• monodepth: Unsupervised Monocular Depth Estimation with Left-Right Consistency [Notes] CVPR 2017 oral (monocular depth estimation, stereo for training)
• Struct2depth: Depth Prediction Without the Sensors: Leveraging Structure for Unsupervised Learning from Monocular Videos [Notes] AAAI 2019 (monocular depth estimation, estimating movement of dynamic object)
• Unsupervised Learning of Geometry with Edge-aware Depth-Normal Consistency [Notes] AAAI 2018 (monocular depth estimation, static assumption, surface normal)
• LEGO Learning Edge with Geometry all at Once by Watching Videos [Notes] CVPR 2018 spotlight (monocular depth estimation, static assumption, surface normal)
• Object Detection and 3D Estimation via an FMCW Radar Using a Fully Convolutional Network [Notes] (radar, RD map, OD, Arxiv 201902)
• A study on Radar Target Detection Based on Deep Neural Networks [Notes] (radar, RD map, OD)
• 2D Car Detection in Radar Data with PointNets [Notes] (from Ulm Univ, radar, point cloud, OD, Arxiv 201904)
• Learning Confidence for Out-of-Distribution Detection in Neural Networks [Notes] (budget to cheat)
• A Deep Learning Approach to Traffic Lights: Detection, Tracking, and Classification [Notes] ICRA 2017 (Bosch, traffic lights)
• How hard can it be? Estimating the difficulty of visual search in an image [Notes] CVPR 2016
• Deep Multi-modal Object Detection and Semantic Segmentation for Autonomous Driving: Datasets, Methods, and Challenges [Notes] (review from Bosch)
• Review of monocular 3d object detection (blog from 知乎)
• Deep3dBox: 3D Bounding Box Estimation Using Deep Learning and Geometry [Notes] CVPR 2017 [Zoox]
• MonoPSR: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction [Notes] CVPR 2019
• OFT: Orthographic Feature Transform for Monocular 3D Object Detection [Notes] BMVC 2019 [Convert camera to BEV, Alex Kendall]

• MixMatch: A Holistic Approach to Semi-Supervised Learning [Notes]
• EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks [Notes] ICML 2019
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision? [Notes] NIPS 2017
• Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understanding [Notes]BMVC 2017
• TrafficPredict: Trajectory Prediction for Heterogeneous Traffic-Agents [Notes] AAAI 2019 oral
• Deep Depth Completion of a Single RGB-D Image [Notes] CVPR 2018 (indoor)
• DeepLiDAR: Deep Surface Normal Guided Depth Prediction for Outdoor Scene from Sparse LiDAR Data and Single Color Image [Notes] CVPR 2019 (outdoor)
• SfMLearner: Unsupervised Learning of Depth and Ego-Motion from Video [Notes] CVPR 2017
• Monodepth2: Digging Into Self-Supervised Monocular Depth Estimation [Notes] ICCV 2019 [Niantic]
• DeepSignals: Predicting Intent of Drivers Through Visual Signals [Notes] ICRA 2019 (@Uber, turn signal detection)
• FCOS: Fully Convolutional One-Stage Object Detection [Notes] ICCV 2019
• Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving [Notes] ICLR 2020
• MMF: Multi-Task Multi-Sensor Fusion for 3D Object Detection [Notes] CVPR 2019 (@Uber, sensor fusion)

• CenterNet: Objects as points (from ExtremeNet authors) [Notes]
• CenterNet: Object Detection with Keypoint Triplets [Notes]
• Object Detection based on Region Decomposition and Assembly [Notes] AAAI 2019
• The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks [Notes] ICLR 2019
• M2Det: A Single-Shot Object Detector based on Multi-Level Feature Pyramid Network [Notes] AAAI 2019
• Deep Radar Detector [Notes] RadarCon 2019
• Semantic Segmentation on Radar Point Clouds [[Notes]] (from Daimler AG) FUSION 2018
• Pruning Filters for Efficient ConvNets [Notes] ICLR 2017
• Layer-compensated Pruning for Resource-constrained Convolutional Neural Networks [Notes] NIPS 2018 talk
• LeGR: Filter Pruning via Learned Global Ranking [Notes]
• NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detection [Notes] CVPR 2019
• AutoAugment: Learning Augmentation Policies from Data [Notes] CVPR 2019
• Path Aggregation Network for Instance Segmentation [Notes] CVPR 2018
• Channel Pruning for Accelerating Very Deep Neural Networks ICCV 2017 (Face++, Yihui He) [Notes]
• AMC: AutoML for Model Compression and Acceleration on Mobile Devices ECCV 2018 (Song Han, Yihui He)
• MobileNetV3: Searching for MobileNetV3 [Notes]
• MnasNet: Platform-Aware Neural Architecture Search for Mobile [Notes] CVPR 2019
• Rethinking the Value of Network Pruning ICLR 2019

• MobileNetV2: Inverted Residuals and Linear Bottlenecks (MobileNets v2) [Notes] CVPR 2018
• A New Performance Measure and Evaluation Benchmark for Road Detection Algorithms [Notes] ITSC 2013
• MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving [Notes]
• Optimizing the Trade-off between Single-Stage and Two-Stage Object Detectors using Image Difficulty Prediction (Very nice illustration of 1 and 2 stage object detection)
• Light-Head R-CNN: In Defense of Two-Stage Object Detector [Notes] (from Megvii)
• CSP: High-level Semantic Feature Detection: A New Perspective for Pedestrian Detection (center and scale prediction) [Notes] CVPR 2019
• Review of Anchor-free methods (知乎Blog) 目标检测：Anchor-Free时代 Anchor free深度学习的目标检测方法 My Slides on CSP
• DenseBox: Unifying Landmark Localization with End to End Object Detection
• CornerNet: Detecting Objects as Paired Keypoints [Notes] ECCV 2018
• ExtremeNet: Bottom-up Object Detection by Grouping Extreme and Center Points [Notes] CVPR 2019
• FSAF: Feature Selective Anchor-Free Module for Single-Shot Object Detection [Notes] CVPR 2019
• FoveaBox: Beyond Anchor-based Object Detector (anchor-free) [Notes]

• Bag of Freebies for Training Object Detection Neural Networks [Notes]
• mixup: Beyond Empirical Risk Minimization [Notes] ICLR 2018
• Multi-view Convolutional Neural Networks for 3D Shape Recognition (MVCNN) [Notes] ICCV 2015
• 3D ShapeNets: A Deep Representation for Volumetric Shapes [Notes] CVPR 2015
• Volumetric and Multi-View CNNs for Object Classification on 3D Data [Notes] CVPR 2016
• Group Normalization [Notes] ECCV 2018
• Spatial Transformer Networks [Notes] NIPS 2015
• Frustum PointNets for 3D Object Detection from RGB-D Data (F-PointNet) [Notes] CVPR 2018
• Dynamic Graph CNN for Learning on Point Clouds [Notes]
• PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud (SOTA for 3D object detection) [Notes] CVPR 2019
• Multi-View 3D Object Detection Network for Autonomous Driving (MV3D) [Notes] CVPR 2017 (Baidu, sensor fusion, BV proposal)
• Joint 3D Proposal Generation and Object Detection from View Aggregation (AVOD) [Notes] IROS 2018 (sensor fusion, multiview proposal)
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications [Notes]
• Pseudo-LiDAR from Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving [Notes] CVPR 2019
• VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection CVPR 2018 (Apple, first end-to-end point cloud encoding to grid)
• SECOND: Sparsely Embedded Convolutional Detection Sensors 2018 (builds on VoxelNet)
• PointPillars: Fast Encoders for Object Detection from Point Clouds [Notes] CVPR 2019 (builds on SECOND)
• Are we ready for Autonomous Driving? The KITTI Vision Benchmark Suite [Notes] CVPR 2012
• Vision meets Robotics: The KITTI Dataset [Notes] IJRR 2013

• Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset (I3D) [Notes]Video CVPR 2017
• Initialization Strategies of Spatio-Temporal Convolutional Neural Networks [Notes] Video
• Detect-and-Track: Efficient Pose Estimation in Videos [Notes] ICCV 2017 Video
• Deep Learning Based Rib Centerline Extraction and Labeling [Notes] MI MICCAI 2018
• SlowFast Networks for Video Recognition [Notes] ICCV 2019 Oral
• Aggregated Residual Transformations for Deep Neural Networks (ResNeXt) [Notes] CVPR 2017
• Beyond the pixel plane: sensing and learning in 3D (blog, 中文版本)
• VoxNet: A 3D Convolutional Neural Network for Real-Time Object Recognition (VoxNet) [Notes]
• PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation CVPR 2017 [Notes]
• PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space NIPS 2017 [Notes]
• Review of Geometric deep learning 几何深度学习前沿 (from 知乎) (Up to CVPR 2018)

• Human-level control through deep reinforcement learning (Nature DQN paper) [Notes] DRL
• Retina U-Net: Embarrassingly Simple Exploitation of Segmentation Supervision for Medical Object Detection [Notes] MI
• Panoptic Segmentation [Notes] PanSeg
• Panoptic Feature Pyramid Networks [Notes] PanSeg
• Attention-guided Unified Network for Panoptic Segmentation [Notes] PanSeg
• Bag of Tricks for Image Classification with Convolutional Neural Networks [Notes] CLS
• Deep Reinforcement Learning for Vessel Centerline Tracing in Multi-modality 3D Volumes [Notes] DRL MI
• Deep Reinforcement Learning for Flappy Bird [Notes] DRL
• Long-Term Feature Banks for Detailed Video Understanding [Notes] Video
• Non-local Neural Networks [Notes] Video CVPR 2018

• Mask R-CNN
• Cascade R-CNN: Delving into High Quality Object Detection
• Focal Loss for Dense Object Detection (RetinaNet) [Notes]
• Squeeze-and-Excitation Networks (SENet)
• Progressive Growing of GANs for Improved Quality, Stability, and Variation
• Deformable Convolutional Networks (build on R-FCN)
• Learning Region Features for Object Detection

• Learning notes on Deep Learning
• List of Papers on Machine Learning
• Notes of Literature Review on CNN in CV This is the notes for all the papers in the recommended list here
• Notes of Literature Review (Others)
• Notes on how to set up DL/ML environment
• Useful setup notes

Here is the list of papers waiting to be read.

• SqueezeDet: Unified, Small, Low Power Fully Convolutional Neural Networks for Real-Time Object Detection for Autonomous Driving
• Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
• ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness ICML 2019
• Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNet (BagNet) blog ICML 2019
• A disciplined approach to neural network hyper-parameters: Part 1 -- learning rate, batch size, momentum, and weight decay
• Understanding deep learning requires rethinking generalization

• Mask Scoring R-CNN CVPR 2019
• Training Region-based Object Detectors with Online Hard Example Mining

• Learning Spatiotemporal Features with 3D Convolutional Networks (C3D) Video ICCV 2015
• AVA: A Video Dataset of Spatio-temporally Localized Atomic Visual Actions
• Spatiotemporal Residual Networks for Video Action Recognition (decouple spatiotemporal) NIPS 2016
• Learning Spatio-Temporal Representation with Pseudo-3D Residual Networks (P3D, decouple spatiotemporal) ICCV 2017
• A Closer Look at Spatiotemporal Convolutions for Action Recognition (decouple spatiotemporal) CVPR 2018
• Rethinking Spatiotemporal Feature Learning: Speed-Accuracy Trade-offs in Video Classification (decouple spatiotemporal) ECCV 2018
• Can Spatiotemporal 3D CNNs Retrace the History of 2D CNNs and ImageNet? CVPR 2018

• Concurrent Spatial and Channel Squeeze & Excitation in Fully Convolutional Networks
• CBAM: Convolutional Block Attention Module

• Playing Atari with Deep Reinforcement Learning NIPS 2013
• Multi-Scale Deep Reinforcement Learning for Real-Time 3D-Landmark Detection in CT Scan
• An Artificial Agent for Robust Image Registration

• 3D-CNN：3D Convolutional Neural Networks for Landing Zone Detection from LiDAR
• Generative and Discriminative Voxel Modeling with Convolutional Neural Networks
• Orientation-boosted Voxel Nets for 3D Object Recognition (ORION) <BMVC 2017>
• GIFT: A Real-time and Scalable 3D Shape Search Engine CVPR 2016
• 3D Shape Segmentation with Projective Convolutional Networks (ShapePFCN)CVPR 2017
• Learning Local Shape Descriptors from Part Correspondences With Multi-view Convolutional Networks
• Open3D: A Modern Library for 3D Data Processing
• Multimodal Deep Learning for Robust RGB-D Object Recognition IROS 2015
• FlowNet3D: Learning Scene Flow in 3D Point Clouds CVPR 2019
• Mining Point Cloud Local Structures by Kernel Correlation and Graph Pooling CVPR 2018 (Neighbors Do Help: Deeply Exploiting Local Structures of Point Clouds)
• PU-Net: Point Cloud Upsampling Network CVPR 2018
• Recurrent Slice Networks for 3D Segmentation of Point Clouds CVPR 2018
• SPLATNet: Sparse Lattice Networks for Point Cloud Processing CVPR 2018
• Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering NIPS 2016
• Semi-Supervised Classification with Graph Convolutional Networks ICLR 2017
• Geometric Matrix Completion with Recurrent Multi-Graph Neural Networks NIPS 2017
• Graph Attention Networks ICLR 2018
• 3D-SSD: Learning Hierarchical Features from RGB-D Images for Amodal 3D Object Detection (3D SSD)
• Escape from Cells: Deep Kd-Networks for the Recognition of 3D Point Cloud Models ICCV 2017
• Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesis CVPR 2017
• IPOD: Intensive Point-based Object Detector for Point Cloud
• Amodal Detection of 3D Objects: Inferring 3D Bounding Boxes from 2D Ones in RGB-Depth Images CVPR 2017
• 2D-Driven 3D Object Detection in RGB-D Images
• 3D-SSD: Learning Hierarchical Features from RGB-D Images for Amodal 3D Object Detection

• Learning Depth with Convolutional Spatial Propagation Network (Baidu, depth from SPN) ECCV 2018
• Model Vulnerability to Distributional Shifts over Image Transformation Sets (CVPR workshop) tl:dr
• A Unified Panoptic Segmentation Network CVPR 2019 PanSeg
• FastDraw: Addressing the Long Tail of Lane Detection by Adapting a Sequential Prediction Network
• PSMNet: Pyramid Stereo Matching Network CVPR 2018
• Stereo R-CNN based 3D Object Detection for Autonomous Driving CVPR 2019
• Deep Rigid Instance Scene Flow CVPR 2019
• GeoNet: Deep Geodesic Networks for Point Cloud Analysis CVPR 2019 (oral, Megvii)
• StixelNet: A Deep Convolutional Network for Obstacle Detection and Road Segmentation
• DenseBox: Unifying Landmark Localization with End to End Object Detection
• Calibration of Heterogeneous Sensor Systems
• nuScenes: A multimodal dataset for autonomous driving (dataset, point cloud, radar)
• Xception: Deep Learning with Depthwise Separable Convolutions (Xception)
• Intro：Sensor Fusion for Adas 无人驾驶中的数据融合 (from 知乎) (Up to CVPR 2018)
• Deep Hough Voting for 3D Object Detection in Point Clouds (from Charles Qi)
• Efficient Deep Learning Inference based on Model Compression (Model Compression)
• FCNN: Fourier Convolutional Neural Networks (FFT as CNN)
• Visualizing the Loss Landscape of Neural Nets NIPS 2018
• Automatic adaptation of object detectors to new domains using self-training CVPR 2019 (find corner case and boost)
• Missing Labels in Object Detection CVPR 2019
• Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics (uncertainty)
• Learning to Drive from Simulation without Real World Labels ICRA 2019 (domain adaptation, sim2real)
• Unsupervised Learning of Depth and Ego-Motion from Monocular Video Using 3D Geometric Constraints CVPR 2018
• YUVMultiNet: Real-time YUV multi-task CNN for autonomous driving CVPR 2019 (Real Time, Low Power)
• Polar Transformer Networks ICLR 2018
• Circular Object Detection in Polar Coordinates for 2D LIDAR Data CCPR 2016
• Adaptive Scheduling for Multi-Task Learning NIPS 2018 (NMT)
• A Novel Approach for Detecting Road Based on Two-Stream Fusion Fully Convolutional Network (convert camera to BEV)
• In-Place Activated BatchNorm for Memory-Optimized Training of DNNs CVPR 2018 (optimized BatchNorm + ReLU)
• Deep Fusion of Heterogeneous Sensor Modalities for the Advancements of ADAS to Autonomous Vehicles
• Revisiting Small Batch Training for Deep Neural Networks
• ICML2019 workshop: Adaptive and Multitask Learning: Algorithms & Systems ICML 2019

• The DriveU Traffic Light Dataset: Introduction and Comparison with Existing Datasets ICRA 2018
• The Oxford Radar RobotCar Dataset: A Radar Extension to the Oxford RobotCar Dataset
• Vision for Looking at Traffic Lights: Issues, Survey, and Perspectives (traffic light survey, UCSD LISA)
• Review of Graph Spectrum Theory (WIP)
• 3D Deep Learning Tutorial at CVPR 2017 [Notes] - (WIP)
• A Survey on Neural Architecture Search
• Network pruning tutorial (blog)
• GNN tutorial at CVPR 2019

• Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentation 3DV 2018
• Depth Map Prediction from a Single Image using a Multi-Scale Deep Network NIPS 2014 (Eigen et al)
• Learning Depth from Monocular Videos using Direct Methods CVPR 2018 (monocular depth estimation)
• Virtual-Normal: Enforcing geometric constraints of virtual normal for depth prediction [Notes] ICCV 2019 (better generation of PL)
• Exploiting temporal consistency for real-time video depth estimation ICCV 2019
• Self-supervised Learning with Geometric Constraints in Monocular Video: Connecting Flow, Depth, and Camera ICCV 2019
• Spatial Correspondence with Generative Adversarial Network: Learning Depth from Monocular Videos ICCV 2019
• Unsupervised Collaborative Learning of Keyframe Detection and Visual Odometry Towards Monocular Deep SLAM ICCV 2019
• Visualization of Convolutional Neural Networks for Monocular Depth Estimation ICCV 2019
• Self-Supervised Monocular Depth Hints ICCV 2019

• PIXOR: Real-time 3D Object Detection from Point Clouds CVPR 2018 (birds eye view)
• PointSIFT: A SIFT-like Network Module for 3D Point Cloud Semantic Segmentation (pointnet alternative, backbone)
• Vehicle Detection from 3D Lidar Using Fully Convolutional Network (VeloFCN) RSS 2016
• KPConv: Flexible and Deformable Convolution for Point Clouds (from the authors of PointNet)
• PointCNN: Convolution On X-Transformed Points NIPS 2018
• L3-Net: Towards Learning based LiDAR Localization for Autonomous Driving CVPR 2019
• RoarNet: A Robust 3D Object Detection based on RegiOn Approximation Refinement (sensor fusion, 3D mono proposal, refined in point cloud)
• DeLS-3D: Deep Localization and Segmentation with a 3D Semantic Map CVPR 2018

• Classification of Objects in Polarimetric Radar Images Using CNNs at 77 GHz (Radar, polar) <-- todo
• Gated2Depth: Real-time Dense Lidar from Gated Images ICCV 2019 oral
• PifPaf: Composite Fields for Human Pose Estimation CVPR 2019
• Uncertainty Guided Multi-Scale Residual Learning-using a Cycle Spinning CNN for Single Image De-Raining CVPR 2019
• Hybrid Task Cascade for Instance Segmentation CVPR 2019 (cascaded mask RCNN)
• 3DOP: 3D Object Proposals for Accurate Object Class Detection NIPS 2015
• DirectShape: Photometric Alignment of Shape Priors for Visual Vehicle Pose and Shape Estimation
• Eliminating the Blind Spot: Adapting 3D Object Detection and Monocular Depth Estimation to 360° Panoramic Imagery ECCV 2018 (Monocular 3D object detection and depth estimation)
• Measuring Calibration in Deep Learning CVPR 2019
• Sampling-free Epistemic Uncertainty Estimation Using Approximated Variance Propagation ICCV 2019 (epistemic uncertainty)
• End-to-end Lane Detection through Differentiable Least-Squares Fitting ICCV 2019
• Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection IROS 2019
• Dropout Sampling for Robust Object Detection in Open-Set Conditions ICRA 2018 (Niko Sünderhauf)
• Evaluating Merging Strategies for Sampling-based Uncertainty Techniques in Object Detection ICRA 2019 (Niko Sünderhauf)
• GCNet: End-to-End Learning of Geometry and Context for Deep Stereo Regression ICCV 2017 (disparity estimation)
• PSMNet: Pyramid Stereo Matching Network CVPR 2018 (disparity estimation)
• Practical Deep Stereo (PDS): Toward applications-friendly deep stereo matching NIPS 2018 (disparity estimation)
• Self-supervised Sparse-to-Dense: Self-supervised Depth Completion from LiDAR and Monocular Camera ICRA 2019/kbd>
• Road Scene Understanding by Occupancy Grid Learning from Sparse Radar Clusters using Semantic Segmentation ICCV 2019 (radar)
• Towards Scene Understanding: Unsupervised Monocular Depth Estimation with Semantic-aware Representation CVPR 2019
• DDP: Dense Depth Posterior from Single Image and Sparse Range CVPR 2019
• End-to-End Learning of Geometry and Context for Deep Stereo Regression ICCV 2017
• Understanding deep learning requires rethinking generalization ICLR 2017 (best paper)
• PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation CVPR 2018 (sensor fusion)
• CNNs for Interference Mitigation and Denoising in Automotive Radar Using Real-World Data NeurIPS 2019 (radar)
• A Multi-Sensor Fusion System for Moving Object Detection and Tracking in Urban Driving Environments ICRA 2014
• Domain Adaptive Faster R-CNN for Object Detection in the Wild CVPR 2018
• Foggy Cityscapes: Semantic Foggy Scene Understanding with Synthetic Data IJCV 2018
• Foggy Cityscapes ECCV: Model Adaptation with Synthetic and Real Data for Semantic Dense Foggy Scene Understanding ECCV 2018
• A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks ICLR 2017 (NLL score as anomaly score)
• Augmented Reality Meets Computer Vision : Efficient Data Generation for Urban Driving Scenes IJCV 2018 (data augmentation with AR, Toyota)

• CRF
• Visual SLAM and Visual Odometry
• ORB SLAM
• ORB (ICCV 2011)
• Bundle Adjustment
• 3D vision
• Codebase of STN
• Codebase of monodepth
• Codebase of KITTI devkit
• SLAM/VIO学习总结