Research map / 3D vision / 3D generation
4DSegStreamer: Streaming 4D Panoptic Segmentation via Dual Threads
Ling Liu, Jun Tian, Li Yi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add memory module (ablation) | P3Former · SemanticKITTI val (unknown pose) | sLSTQ: 0.304 -> 0.349 | Improved |
| add ego-pose forecasting (ablation) | P3+Mem · SemanticKITTI val (unknown pose) | sLSTQ: 0.349 -> 0.497 | Improved |
| add flow forecasting (ablation) | P3+Mem+Pose · SemanticKITTI val (unknown pose) | sLSTQ: 0.497 -> 0.591 | Improved |
| add moving mask so only moving objects get non-zero flow (ablation) | P3+Mem+Pose+Flow · SemanticKITTI val (unknown pose) | sLSTQ: 0.591 -> 0.613 | Improved |
| add flow forecasting with ground-truth pose (ablation) | P3+Mem+GTpose · SemanticKITTI val (known pose) | sLSTQ: 0.563 -> 0.655 | Improved |
| use inverse flow iteration instead of backward flow (ablation) | P3Former · SemanticKITTI val (unknown pose) | sLSTQ: 0.565 -> 0.613 | Improved |
| use inverse flow iteration instead of forward flow (ablation) | P3Former · SemanticKITTI val (unknown pose) | sLSTQ: 0.589 -> 0.613 | Improved |
| use inverse flow iteration instead of inverse brute search (ablation) | P3Former · SemanticKITTI val (unknown pose) | sLSTQ: 0.591 -> 0.613 | Improved |
| add 4DSegStreamer to Mask4Former | Mask4Former · SemanticKITTI val (unknown pose) | sLSTQ: 0.515 -> 0.688 | Improved |
| add 4DSegStreamer to Eq-4D-StOP | Eq-4D-StOP · SemanticKITTI val (unknown pose) | sLSTQ: 0.504 -> 0.674 | Improved |
| add 4DSegStreamer to P3former | P3former · SemanticKITTI val (unknown pose) | sLSTQ: 0.304 -> 0.613 | Improved |
| run on A100 instead of 3090 | 4DSegStreamer (M4F) · SemanticKITTI val | sLSTQ: 0.688 -> 0.702 | Improved |
Closest papers by meaning
- IGGT4D: Streaming 4D Instance-Grounded Geometry Transformer (NeurIPS 2026 · similarity 0.88)
- Motion4D: Learning 3D-Consistent Motion and Semantics for 4D Scene Understanding (NeurIPS 2025 · similarity 0.86)
- Flux4D: Flow-based Unsupervised 4D Reconstruction (NeurIPS 2025 · similarity 0.86)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 2025 · similarity 0.85)
- PanoRecon: Real-Time Panoptic 3D Reconstruction from Monocular Video (CVPR 2024 · similarity 0.85)
- Real-time Stereo-based 3D Object Detection for Streaming Perception (NeurIPS 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
