占有予測/追跡2026/3/1arXiv:2603.08521

OccTrack360: 全方位魚眼カメラによる4Dパノプティック占有追跡

OccTrack360: 4D Panoptic Occupancy Tracking from Surround-View Fisheye Cameras

全方位魚眼カメラ向けの長尺4Dパノプティック占有追跡ベンチマークOccTrack360と、球面歪みと位置ずれに対処するベースライン手法FoSOccを提案した。

詳しい要約

1. どんなもの？

- 4D panoptic occupancy trackingの新benchmark **OccTrack360**を提案。 - surround-view fisheye camerasからの4D panoptic occupancy trackingを対象。 - 174~2234 framesの長く多様なsequenceを提供。 - voxel visibility annotationとしてall-direction occlusion maskとMEI-based fisheye field-of-view maskを付与。 - fisheye向けbaseline **FoSOcc**も提案。

2. 先行研究と比べてどこがすごい？

- 従来はsurround-view fisheye sensing、long temporal sequence、instance-level voxel trackingを支えるbenchmarkが不足。 - OccTrack360は既存benchmarkより大幅に長く多様なsequence(174~2234 frames)を提供。 - principled voxel visibility annotationを備える点が新しい。 - FoSOccはfisheye特有の歪みとvoxel-space localizationの課題に対処。 - Occ3D-WaymoとOccTrack360でoccupancy tracking品質を改善。

3. 技術・手法の肝は？

- **FoSOcc**はfisheye occupancy trackingの2課題(distorted spherical projection、inaccurate voxel-space localization)に対処。 - **Center Focusing Module (CFM)**: supervised focus guidanceによりinstance-aware spatial localizationを強化。 - **Fisheye-based Enhanced Lifting (FEL)**: Unified Projection Modelの下でperspective liftingをfisheye imagingへ拡張。 - これらによりfisheye向けの強いbaselineを構築。

4. どうやって有効だと検証した？

- **Occ3D-Waymo**と**OccTrack360**でextensive experimentsを実施。 - occupancy tracking品質が向上し、特にgeometrically regular categoriesで顕著なgain。 - surround-view fisheye 4D occupancy trackingのstrong baselineを確立。 - benchmarkとsource codeは公開予定(https://github.com/YouthZest-Lin/OccTrack360)。

5. 議論はある？

- 4D panoptic occupancy trackingのbenchmark不足が研究進展の制限要因と指摘。 - fisheye sensing、long temporal sequence、instance-level voxel trackingの重要性を強調。 - geometrically regular categoriesでの改善が顕著である一方、他のカテゴリの詳細は要旨からは不明。 - 今後のsurround-view fisheye 4D occupancy tracking研究の基盤を提供。

6. 次に読むべき論文は？

- **Occ3D-Waymo** (比較対象benchmark) - **OccTrack360** (本論文のbenchmark) - **FoSOcc** (本論文のbaseline) - **Unified Projection Model** (FELの基盤) - **MEI-based fisheye field-of-view mask** (関連手法) - 同分野の定番としてoccupancy prediction、4D panoptic occupancy tracking、surround-view fisheye sensingの関連研究。

※ AIが要旨から生成した要約です。正確性は原文をご確認ください。

著者: Yongzhi Lin, Kai Luo, Yuanfan Zheng, Hao Shi, Mengfei Duan, Yang Liu, Kailun Yang

分類: cs.CV, cs.RO, eess.IV

原文アブストラクト

Understanding dynamic 3D environments in a spatially continuous and temporally consistent manner is fundamental for robotics and autonomous driving. While recent advances in occupancy prediction provide a unified representation of scene geometry and semantics, progress in 4D panoptic occupancy tracking remains limited by the lack of benchmarks that support surround-view fisheye sensing, long temporal sequences, and instance-level voxel tracking. To address this gap, we present OccTrack360, a new benchmark for 4D panoptic occupancy tracking from surround-view fisheye cameras. OccTrack360 provides substantially longer and more diverse sequences (174~2234 frames) than prior benchmarks, together with principled voxel visibility annotations, including an all-direction occlusion mask and an MEI-based fisheye field-of-view mask. To establish a strong fisheye-oriented baseline, we further propose Focus on Sphere Occ (FoSOcc), a framework that addresses two core challenges in fisheye occupancy tracking: distorted spherical projection and inaccurate voxel-space localization. FoSOcc includes a Center Focusing Module (CFM) to enhance instance-aware spatial localization through supervised focus guidance, and a Fisheye-based Enhanced Lifting (FEL) that extends perspective lifting to fisheye imaging under the Unified Projection Model. Extensive experiments on Occ3D-Waymo and OccTrack360 show that our method improves occupancy tracking quality with notable gains on geometrically regular categories, and establishes a strong baseline for future research on surround-view fisheye 4D occupancy tracking. The benchmark and source code will be made publicly available at https://github.com/YouthZest-Lin/OccTrack360.

arXivで原文を読む →arXiv概要ページ

← 論文一覧