从近100篇综述出发，一览自动驾驶的技术发展路线！

点击下方卡片，关注“自动驾驶之心”公众号

ADAS巨卷干货，即可获取

今天自动驾驶之心为大家分享近百篇综述，一览自动驾驶的技术发展路线！如果您有相关工作需要分享，请在文末联系我们！

>>点击进入→自动驾驶之心【全栈算法】技术交流群  

2023年快要过去了，不得不说，今年的技术变更实在很快，在线高精地图、大模型、端到端自动驾驶、世界模型、Occ、Nerf这些新兴技术，慢慢走向量产的计划中，今天自动驾驶Daily就为大家盘下近百篇综述和经典论文，涉及感知、定位、融合、Occupancy、大模型、端到端、规划控制、BEV感知、数据相关等，一览自动驾驶发展路线。

所有论文出自--国内首个自动驾驶学习社区：自动驾驶之心知识星球（点我有惊喜），所有论文均可在星球内下载，更有30+学习路线，近2300人一起讨论。

端到端自动驾驶

1. Recent Advancements in End-to-End Autonomous Driving using Deep Learning: A Survey

2. End-to-end Autonomous Driving: Challenges and Frontiers

在线高精地图

1. HDMapNet：基于语义分割的在线局部高精地图构建 (ICRA2022)

2. VectorMapNet：基于自回归方式的端到端矢量化地图构建（ICML2023）

3. MapTR ： 基于固定数目点的矢量化地图构建 (ICLR2023)

4. MapTRv2：一种在线矢量化高清地图构建的端到端框架

5. PivotNet：基于动态枢纽点的矢量化地图构建 （ICCV2023）

6. BeMapNet：基于贝塞尔曲线的矢量化地图构建 (CVPR2023)

7. LATR:  无显式BEV 特征的3D车道线检测 (ICCV2023)

8. TopoNet: 基于图的驾驶场景拓扑推理

9. TopoMLP: 先检测后推理（拓扑推理 strong pipeline）

10. LaneGAP：连续性在线车道图构建

11. Neural Map Prior: 神经地图先验辅助在线建图 （CVPR2023）

12. MapEX：现有地图先验显著提升在线建图性能

大模型与自动驾驶

1. CLIP：Learning Transferable Visual Models From Natural Language Supervision

2. BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation

3. BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models

4. InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning

5. MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models

6. InstructGPT：Training language models to follow instructions with human feedback

7. ADAPT: Action-aware Driving Caption Transformer

8. BEVGPT：Generative Pre-trained Large Model for Autonomous Driving Prediction, Decision-Making, and Planning

9. DriveGPT4：Interpretable End-to-end Autonomous Driving via Large Language Model

10. Drive Like a Human Rethinking Autonomous Driving with Large Language Models

11. Driving with LLMs: Fusing Object-Level Vector Modality for Explainable Autonomous Driving

12. HiLM-D: Towards High-Resolution Understanding in Multimodal Large Language Models for Autonomous Driving

13. LanguageMPC: Large Language Models as Decision Makers for Autonomous Driving

14. Planning-oriented Autonomous Driving

15. WEDGE A multi-weather autonomous driving dataset built from generative vision-language models

Nerf与自动驾驶

1. NeRF: Neural Radiance Field in 3D Vision, A Comprehensive Review

2. Neural Volume Rendering: NeRF And Beyond

3. MobileNeRF：移动端实时渲染，Nerf导出Mesh（CVPR2023）

4. Co-SLAM：实时视觉定位和NeRF建图（CVPR2023）

5. Neuralangelo：当前最好的NeRF表面重建方法（CVPR2023）

6. MARS：首个开源自动驾驶NeRF仿真工具（CICAI2023）

7. UniOcc：NeRF和3D占用网络（AD2023 Challenge）

8. Unisim：自动驾驶场景的传感器模拟（CVPR2023）

Occupancy占用网络

1. Grid-Centric Traffic Scenario Perception for Autonomous Driving: A Comprehensive Review

BEV感知

1. Vision-Centric BEV Perception: A Survey

2. Vision-RADAR fusion for Robotics BEV Detections: A Survey

3. Surround-View Vision-based 3D Detection for Autonomous Driving: A Survey

4. Delving into the Devils of Bird’s-eye-view Perception: A Review, Evaluation and Recipe

多模态融合

针对Lidar、Radar、视觉等数据方案进行融合感知；

1. A Survey on Deep Domain Adaptation for LiDAR Perception

2. Automatic Target Recognition on Synthetic Aperture Radar Imagery：A Survey

3. Deep Multi-modal Object Detection and Semantic Segmentation for Autonomous Driving：Datasets, Methods, and Challenges

4. MmWave Radar and Vision Fusion for Object Detection in Autonomous Driving：A Review

5. Multi-Modal 3D Object Detection in Autonomous Driving：A Survey

6. Multi-modal Sensor Fusion for Auto Driving Perception：A Survey

7. Multi-Sensor 3D Object Box Refinement for Autonomous Driving

8. Multi-View Fusion of Sensor Data for Improved Perception and Prediction in Autonomous Driving

3D检测

对基于单目图像、双目图像、点云数据、多模态数据的3D检测方法进行了梳理；

1. 3D Object Detection for Autonomous Driving：A Review and New Outlooks

2. 3D Object Detection from Images for Autonomous Driving A Survey

3. A Survey of Robust LiDAR-based 3D Object Detection Methods for autonomous driving

4. A Survey on 3D Object Detection Methods for Autonomous Driving Applications

5. Deep Learning for 3D Point Cloud Understanding：A Survey

6. Multi-Modal 3D Object Detection in Autonomous Driving：a survey

7. Survey and Systematization of 3D Object Detection Models and Methods

目标检测综述

主要涉及通用目标检测任务、检测任务中的数据不均衡问题、伪装目标检测、自动驾驶领域检测任务、anchor-based、anchor-free、one-stage、two-stage方案等；

1. A Survey of Deep Learning for Low-Shot Object Detection

2. A Survey of Deep Learning-based Object Detection

3. Camouflaged Object Detection and Tracking：A Survey

4. Deep Learning for Generic Object Detection：A Survey

5. Imbalance Problems in Object Detection：A survey

6. Object Detection in 20 Years：A Survey

7. Object Detection in Autonomous Vehicles：Status and Open Challenges

8. Recent Advances in Deep Learning for Object Detection

目标检测数据增强与不均衡问题

主要涉及目标检测任务中的数据增强、小目标检测、小样本学习、autoargument等工作；

1. A survey on Image Data Augmentation for Deep Learning

2. Augmentation for small object detection

3. Bag of Freebies for Training Object Detection Neural Networks

4. Generalizing from a Few Examples：A Survey on Few-Shot

5. Learning Data Augmentation Strategies for Object Detection

分割综述

主要对实时图像分割、视频分割、实例分割、弱监督/无监督分割、点云分割等方案展开讨论；

1. A Review of Point Cloud Semantic Segmentation

2. A SURVEY ON DEEP LEARNING METHODS FOR SEMANTIC IMAGE SEGMENTATION IN REAL-TIME

3. A SURVEY ON DEEP LEARNING METHODS FOR SEMANTIC

4. A Survey on Deep Learning Technique for Video Segmentation

5. A Survey on Instance Segmentation State of the art

6. A Survey on Label-efficient Deep Segmentation-Bridging the Gap between Weak Supervision and Dense Prediction

7. A Technical Survey and Evaluation of Traditional Point Cloud Clustering  for LiDAR Panoptic Segmentation

8. Evolution of Image Segmentation using Deep Convolutional Neural Network A Survey

9. On Efficient Real-Time Semantic Segmentation

10. Unsupervised Domain Adaptation for Semantic Image Segmentation-a Comprehensive Survey

多任务学习

对检测+分割+关键点+车道线联合任务训练方法进行了汇总；

1. Cascade R-CNN

2. Deep Multi-Task Learning for Joint Localization, Perception, and Prediction

3. Mask R-CNN

4. Mask Scoring R-CNN

5. Multi-Task Multi-Sensor Fusion for 3D Object Detection

6. MultiTask-CenterNet

7. OmniDet

8. YOLOP

9. YOLO-Pose

目标跟踪

对单目标和多目标跟踪、滤波和端到端方法进行了汇总；

1. Camouflaged Object Detection and Tracking：A Survey

2. Deep Learning for UAV-based Object Detection and Tracking：A Survey

3. Deep Learning on Monocular Object Pose Detection and Tracking：A Comprehensive Overview

4. Detection, Recognition, and Tracking：A Survey

5. Infrastructure-Based Object Detection and Tracking for Cooperative Driving Automation：A Survey

6. Recent Advances in Embedding Methods for Multi-Object Tracking：A Survey

7. Single Object Tracking：A Survey of Methods, Datasets, and Evaluation Metrics

8. Visual Object Tracking with Discriminative Filters and Siamese Networks：A Survey and Outlook

深度估计

针对单目、双目深度估计方法进行了汇总，对户外常见问题与精度损失展开了讨论；

1. A Survey on Deep Learning Techniques for Stereo-based Depth Estimation

2. Deep Learning based Monocular Depth Prediction：Datasets, Methods and Applications

3. Monocular Depth Estimation Based On Deep Learning：An Overview

4. Monocular Depth Estimation：A Survey

5. Outdoor Monocular Depth Estimation：A Research Review

6. Towards Real-Time Monocular Depth Estimation for Robotics：A Survey

关键点检测

人体关键点检测方法汇总，对车辆关键点检测具有一定参考价值；

1. 2D Human Pose Estimation：A Survey

2. A survey of top-down approaches for human pose estimation

3. Efficient Annotation and Learning for 3D Hand Pose Estimation：A Survey

4. Recent Advances in Monocular 2D and 3D Human Pose Estimation：A Deep Learning Perspective

Transformer综述

视觉transformer、轻量级transformer方法汇总；

1. A Survey of Visual Transformers

2. A Survey on Visual Transformer

3. Efficient Transformers：A Survey

车道线检测

对2D/3D车道线检测方法进行了汇总，基于分类、检测、分割、曲线拟合等；

2D车道线

1. A Keypoint-based Global Association Network for Lane Detection

2. CLRNet：Cross Layer Refinement Network for Lane Detection

3. End-to-End Deep Learning of Lane Detection and Path Prediction for Real-Time Autonomous Driving

4. End-to-end Lane Detection through Differentiable Least-Squares Fitting

5. Keep your Eyes on the Lane：Real-time Attention-guided Lane Detection

6. LaneNet：Real-Time Lane Detection Networks for Autonomous Driving

7. Towards End-to-End Lane Detection：an Instance Segmentation Approach

8. Ultra Fast Structure-aware Deep Lane Detection

3D车道线

1. 3D-LaneNet+：Anchor Free Lane Detection using a Semi-Local Representation

2. Deep Multi-Sensor Lane Detection

3. FusionLane：Multi-Sensor Fusion for Lane Marking Semantic Segmentation Using Deep Neural Networks

4. Gen-LaneNet：A Generalized and Scalable Approach for 3D Lane Detection

5. ONCE-3DLanes：Building Monocular 3D Lane Detection

6. 3D-LaneNet：End-to-End 3D Multiple Lane Detection

SLAM综述

定位与建图方案汇总；

1. A Survey on Active Simultaneous Localization and Mapping-State of the Art and New Frontiers

2. The Revisiting Problem in Simultaneous Localization and Mapping-A Survey on Visual Loop Closure Detection

3. From SLAM to Situational Awareness-Challenges

4. Simultaneous Localization and Mapping Related Datasets-A Comprehensive Survey

模型量化

1. A Survey on Deep Neural Network CompressionChallenges, Overview, and Solutions

2. Pruning and Quantization for Deep Neural Network Acceleration A Survey

   

① 全网独家视频课程

BEV感知、毫米波雷达视觉融合、多传感器标定、多传感器融合、多模态3D目标检测、车道线检测、轨迹预测、在线高精地图、世界模型、点云3D目标检测、目标跟踪、Occupancy、cuda与TensorRT模型部署、大模型与自动驾驶、Nerf、语义分割、自动驾驶仿真、传感器部署、决策规划、轨迹预测等多个方向学习视频（扫码即可学习）

视频官网：www.zdjszx.com

② 国内首个自动驾驶学习社区

近2400人的交流社区，涉及30+自动驾驶技术栈学习路线，想要了解更多自动驾驶感知（2D检测、分割、2D/3D车道线、BEV感知、3D目标检测、Occupancy、多传感器融合、多传感器标定、目标跟踪、光流估计）、自动驾驶定位建图（SLAM、高精地图、局部在线地图）、自动驾驶规划控制/轨迹预测等领域技术方案、AI模型部署落地实战、行业动态、岗位发布，欢迎扫描下方二维码，加入自动驾驶之心知识星球，这是一个真正有干货的地方，与领域大佬交流入门、学习、工作、跳槽上的各类难题，日常分享论文+代码+视频，期待交流！

③【自动驾驶之心】技术交流群

自动驾驶之心是首个自动驾驶开发者社区，聚焦目标检测、语义分割、全景分割、实例分割、关键点检测、车道线、目标跟踪、3D目标检测、BEV感知、多模态感知、Occupancy、多传感器融合、transformer、大模型、点云处理、端到端自动驾驶、SLAM、光流估计、深度估计、轨迹预测、高精地图、NeRF、规划控制、模型部署落地、自动驾驶仿真测试、产品经理、硬件配置、AI求职交流等方向。扫码添加汽车人助理微信邀请入群，备注：学校/公司+方向+昵称（快速入群方式）

④【自动驾驶之心】平台矩阵，欢迎联系我们！