GitHub - ryanbgriffiths/ICRA2024PaperList: ICRA2024 Paper List
| Name of the person | Paper title | Paper description in short or comments | Paper link (make sure that it is accesible) |
|---|---|---|---|
| Soumo Roy | ATPPNet: Attention based Temporal Point cloud Prediction Network | Kaustab Pal paper suggested by sir | https://arxiv.org/pdf/2401.17399 |
| Soumo Roy | SLIM: Self-Supervised LiDAR Scene Flow and Motion Segmentation | Scene flow + motion segmentation | https://openaccess.thecvf.com/content/ICCV2021/papers/Baur_SLIM_Self-Supervised_LiDAR_Scene_Flow_and_Motion_Segmentation_ICCV_2021_paper.pdf |
| Soumo Roy | Moving Object Segmentation in 3D LiDAR Data: | ||
| A Learning-based Approach Exploiting Sequential Data | moving object segemtation with a CNN helps to differentiate btw moving and static objects | https://arxiv.org/pdf/2105.08971 | |
| Soumo Roy | InsMOS: Instance-Aware Moving Object Segmentation in LiDAR Data | motion object segmentattion+ spatial temporal information+ instance information | https://arxiv.org/pdf/2105.08971 |
| Soumo Roy | Dynamic 3D Scene Analysis by | ||
| Point Cloud Accumulation | motion object segmentattion + instance information + rigid scene flow estimation | https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136980658.pdf | |
| Soumo Roy | MambaMOS: LiDAR-based 3D Moving Object Segmentation with | ||
| Motion-aware State Space Model | motion object segmentation + Time Clue Bootstrapping Embedding (spatial temporal improvement)+ Motion- | ||
| aware State Space Model (extracting single scan and multiscan features) | https://arxiv.org/pdf/2404.12794 | ||
| Soumo Roy | Moving event detection from LiDAR point streams | M-Detector and tracks point by point uses same hardware (Livox) | https://www.nature.com/articles/s41467-023-44554-8 |
| Soumo Roy | Efficient Moving Object Segmentation in | ||
| LiDAR Point Clouds Using Minimal | |||
| Number of Sweeps | multimodal learning model | ||
| and training on LiDAR point clouds and camera images (using lesser frames for computation imporvement) | https://eprints.sztaki.hu/10870/1/RozsaZ_118_35740965_ny.pdf | ||
| Soumo Roy | An efficient image-guided-based 3D point cloud moving object segmentation with transformer-attention in autonomous driving | MOS + Camera + LIDAR + uses transformer attention based fusion module to fuse features detected | https://www.sciencedirect.com/science/article/pii/S1569843223003126 |
| Soumo Roy | MotionBEV: Attention-Aware Online LiDAR | ||
| Moving Object Segmentation with Bird’s Eye View based Appearance and Motion Features | MOS + spatial temporal info + mutlimodal feature fusion using attention based module + used LIVOX similar setup + tried to validate in a highly dynamic environment like classrooms | https://arxiv.org/pdf/2305.07336 | |
| Soumo Roy | MOVES: Movable and moving LiDAR scene segmentation in label-free settings using static reconstruction | MOS + GAN based adversarial model that segments out moving as well as movable objects in the absence of segmentation information | https://www.sciencedirect.com/science/article/pii/S0031320324004023 |
| Soumo Roy | Receding Moving Object Segmentation | ||
| in 3D LiDAR Data Using Sparse 4D Convolutions | MOS + 4D convolutions + use a binary Bayes filter to recursively integrate | ||
| new predictions of a scan | http://www.ipb.uni-bonn.de/pdfs/mersch2022ral.pdf | ||
| Soumo Roy | LiDAR-SGMOS: Semantics-Guided Moving Object Segmentation with 3D LiDAR | MOS + semantic segmentation with LIDAR range images + cross scan fusion to put LIDAR range images with transformed features | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10341426 |
| Soumo Roy | Robust Moving Objects Detection in Lidar Data | ||
| Exploiting Visual Cues | MOS based on dempster shafer theory + has a novel image validation step to remove false positive + talks about ground plane removal for improving speed | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7759185 | |
| Soumo Roy | UNION: Unsupervised 3D Object Detection | ||
| using Object Appearance-based Pseudo-Classes | step 1, RANSAC is used for ground removal and HDBSCAN is used for spatial clustering | ||
| Step 2 uses ICP-Flow to get motion estimates | |||
| Lastly, step 3 uses DINOv2 for encoding the camera images | https://openreview.net/pdf?id=93gz2lmFtm | ||
| Soumo Roy | Unsupervised Object Detection with LiDAR Clues | 3D pointcloud based detection by taking out lidar clues + repojection on camera based 2D images with YOLO and Faster RCNN + pseudo labelling process | https://openaccess.thecvf.com/content/CVPR2021/papers/Tian_Unsupervised_Object_Detection_With_LIDAR_Clues_CVPR_2021_paper.pdf |
| Soumo Roy | OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point Clouds | the object segmentation network to directly estimate | |
| multi-object masks from a single point cloud frame (using PointNet++ and transformers) + the auxiliary self-supervised | |||
| scene flow estimator (FlowStep 3D) | https://papers.nips.cc/paper_files/paper/2022/file/c6e3856954d23bec921f2d13d8c0e0e7-Paper-Conference.pdf | ||
| Soumo Roy | Zero-Shot 4D Lidar Panoptic Segmentation | pseudo labeling engine + language query on output | https://arxiv.org/pdf/2504.00848 |
| Soumo Roy | LISO: Lidar-only Self-Supervised 3D Object | ||
| Detection | 3D Lidar based scene flow + ego motion estimate (KISS-ICP) + pseudo ground truth | https://www.cvlibs.net/publications/Baur2024ECCV.pdf | |
| Soumo Roy | 3D Object Detection with a Self-supervised | ||
| Lidar Scene Flow Backbone | 3D Lidar based self supervised flow estimation + supervised 3D object detection head | https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136700244.pdf | |
| Soumo Roy | Shelf-Supervised Cross-Modal Pre-Training for | ||
| 3D Object Detection | text promt + query based + 2D to 3D projection + pseudo label + SAM (instance segmentation mask) + Detic(vision language detector) | https://mllmav.github.io/papers/Shelf-Supervised Cross-Modal Pre-Training for 3D Object Detection.pdf | |
Problems observed