└── README.md /README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 |

5 |

Awesome-Thermal-Deep-3D-Learning

6 |

A curated list of resources for deep 3D vision from thermal image 7 |

8 |

9 | 10 |

Table of contents

11 | 12 | - [1. Dataset for Thermal 3D Vision](#1-dataset-for-thermal-3d-vision) 13 | - [2. Deep 3D Vision from Thermal Image](#2-deep-3d-vision-from-thermal-image) 14 | - [2.1 Supervised 3D Vision](#21-supervised-3d-vision-from-thermal-image) 15 | - [2.2 Self-supervised 3D Vision](#22-self-supervised-3d-vision-from-thermal-image) 16 | - [3. Deep 3D Vision from Thermal and Additional Sensor Fusion](#3-deep-3d-vision-from-thermal-and-additional-sensor-fusion) 17 | - [4. Non-Deep 3D Vision from Thermal and Other Sensor](#4-non-deep-3d-vision-from-thermal-and-other-sensor) 18 | 19 | # 1. Dataset for Thermal 3D Vision 20 | 21 |

Dataset List

22 | 23 | |Publication|Title|Sensor|Platform|Scene Types|Dataset Link|Highlight| 24 | |-|-|-|-|-|-|-| 25 | CVPR 17 | [CATS: A Color And Thermal Stereo Benchmark](https://openaccess.thecvf.com/content_cvpr_2017/papers/Treible_CATS_A_Color_CVPR_2017_paper.pdf) | Stereo RGB, Stereo Thermal, Lidar | Handheld | In+Outdoor | [Here](http://bigdatavision.org/cats/)| | 26 | ITS 19 | [KAIST Multi-Spectral DayNight Data Set for Autonomous and Assisted Driving](https://ieeexplore.ieee.org/document/8293689) | Co-aligned RGB/Thermal, Stereo RGB, Lidar, GPS/IMU | Vehicle | Outdoor | Here | Large-scale outdoor dataset | 27 | ICRA 21 | [A Multi-spectral Dataset for Evaluating Motion Estimation Systems](https://arxiv.org/abs/2007.00622) | RGB, Thermal, Kinect, IMU | Handheld | In+Outdoor | [Here](https://github.com/NGCLAB/multi-spectral-dataset)| Large-scale indoor datase| 28 | RA-L 22 | [ViViD++: Vision for Visibility Dataset](https://arxiv.org/abs/2204.06183) | RGB-D, Thermal, Lidar, GPS/IMU, Event | Handheld, Vehicle | In+Outdoor| [Here](https://visibilitydataset.github.io/) | Large-scale outdoor dataset | 29 | Arxiv 22 | [OdomBeyondVision; An Indoor Multi-modal Multi-platform Odometry Dataset Beyond the Visible Spectrum](https://arxiv.org/abs/2206.01589) | RGB, Thermal, Lidar, IMU, Radar | Handheld, UGV, UAV | Indoor | [Here](https://github.com/MAPS-Lab/OdomBeyondVision) | Large-scale indoor dataset| 30 | --- 31 | 32 | 33 | # 2. Deep 3D Vision from Thermal Image 34 | 35 | ## 2.1 Supervised 3D Vision from Thermal Image 36 | 37 |

Paper List

38 | 39 | |Publication|Title|Task|Dataset|Scene Types|Code|Highlight| 40 | |-|-|-|-|-|-|-| 41 | 42 | --- 43 | ## 2.2 Self-Supervised 3D Vision from Thermal Image 44 | 45 |

Paper List

46 | 47 | |Publication|Title|Task|Dataset|Scene Types|Code|Highlight| 48 | |-|-|-|-|-|-|-| 49 | AAAI 18 | [Multispectral Transfer Network: Unsupervised Depth Estimation for All-Day Vision](https://ojs.aaai.org/index.php/AAAI/article/view/12297) | MonoDepth | KAIST Multi-Spectral Dataset | Outdoor | Here | Spatial RGB, thermal image reconstruction | 50 | WACV 21 | [An Alternative of LiDAR in Nighttime: Unsupervised Depth Estimation Based on Single Thermal Image](https://openaccess.thecvf.com/content/WACV2021/papers/Lu_An_Alternative_of_LIDAR_in_Nighttime_Unsupervised_Depth_Estimation_Based_WACV_2021_paper.pdf) | MonoDepth | Own dataset | In+Outdoor | here | Spatial thermal image reconstruction | 51 | RA-L 21 | [SuperThermal: Matching Thermal as Visible Through Thermal Feature Exploration](https://ieeexplore.ieee.org/abstract/document/9359356) | Feature Matching | CSS, KAIST pedestrian dataset | Outdoor | Here | TBA | 52 | RA-L 22& ICRA 22 | [Self-Supervised Depth and Ego-Motion Estimation for Monocular Thermal Video Using Multi-Spectral Consistency Loss](https://ieeexplore.ieee.org/abstract/document/9662239) | MonoDepth, Pose | [ViViD dataset](https://visibilitydataset.github.io/) | In+Outdoor | [Here](https://github.com/UkcheolShin/ThermalSfMLearner-MS) | Temporal RGB, thermal image reconstruction loss | 53 | RA-L 22& IROS 22 | [Maximizing Self-supervision from Thermal Image for Effective Self-supervised Learning of Depth and Ego-motion](https://arxiv.org/abs/2201.04387) | MonoDepth, Pose | [ViViD dataset](https://visibilitydataset.github.io/) | In+Outdoor | [Here](https://github.com/UkcheolShin/ThermalMonoDepth) | Temporal thermal image reconstruction only | 54 | --- 55 | 56 | 57 | 58 | # 3. Deep 3D Vision from Thermal and Additional Sensor Fusion 59 | 60 |

Paper List

61 | 62 | |Publication|Title|Task|Sensor|Dataset|Scene Types|Code|Highlight| 63 | |-|-|-|-|-|-|-|-| 64 | RA-L 20 | [DeepTIO: A Deep Thermal-Inertial Odometry With Visual Hallucination](https://ieeexplore.ieee.org/abstract/document/8968430) | Odometry | Thermal, IMU | Own dataset | Here | TBA | TBA | 65 | IROS 20 | [Tp-tio: A robust thermal-inertial odometry with deep thermalpoint](https://ieeexplore.ieee.org/abstract/document/9341716) | Odometry | Thermal, IMU | TBA | TBA | TBA | TBA | 66 | TRO 21 | [Graph-Based Thermal–Inertial SLAM With Probabilistic Neural Networks](https://ieeexplore.ieee.org/abstract/document/9623261)| Odometry | Thermal, IMU | TBA | TBA | TBA | TBA | 67 | 68 | --- 69 | 70 | # 4. Non-Deep 3D Vision from Thermal and Other Sensor 71 | 72 |

Paper List

73 | 74 | |Publication|Title|Task|Sensor|Dataset|Scene Types|Code|Highlight| 75 | |-|-|-|-|-|-|-|-| 76 | IROS 21 | [Radar Visual Inertial Odometry and Radar Thermal Inertial Odometry: Robust Navigation even in Challenging Visual Conditions](https://ieeexplore.ieee.org/abstract/document/9636799)| Odometry | Thermal, IMU, Radar | TBA | TBA | TBA | TBA | 77 | CVPR 21 | [Shape from Thermal Radiation:Passive Ranging Using Multi-spectral LWIR Measurements](https://openaccess.thecvf.com/content/CVPR2022/html/Nagase_Shape_From_Thermal_Radiation_Passive_Ranging_Using_Multi-Spectral_LWIR_Measurements_CVPR_2022_paper.html) | MonoDepth | Thermal | Own dataset | In+Outdoor | Here | TBA | 78 | RA-L 22 | [Thermal-Inertial SLAM for the Environments With Challenging Illumination](https://ieeexplore.ieee.org/abstract/document/9804793) | Odometry | Thermal, IMU | TBA | TBA | TBA | TBA | 79 | --- 80 | --------------------------------------------------------------------------------