LIST3R: Long-sequence Instance-aware 3D Reconstruction

2026-06-30

AI Summary

Long video sequences often suffer from fragmented and globally inconsistent 3D reconstructions due to viewpoint changes and occlusions. This work proposes a novel paradigm centered on persistent instance anchors: the input video is partitioned into overlapping subsequences, each associated with a local instance database built around anchors that integrate semantic and geometric information. Cross-subsequence anchor matching enables fragment alignment and drives the progressive evolution of a global instance database. By facilitating object-aware fragment association, the method significantly improves camera trajectory accuracy and reconstruction quality on long-sequence benchmarks, demonstrating the critical role of instance anchors in preserving long-term 3D consistency.

Abstract

We present LIST3R, an instance-aware framework for long-sequence 3D reconstruction inspired by the way humans organize spatial memory around stable and recognizable objects. LIST3R organizes long-sequence reconstruction around instance anchors, using them to reconnect fragmented subsequences and consolidate local observations into a coherent global 3D scene. Given a long video, our approach partitions it into overlapping subsequences and builds a structured local instance library for each partial reconstruction, maintaining persistent trackable anchors with semantic and geometric evidence. These anchors are matched across subsequences to recover revisited regions and provide object-aware constraints for fragment alignment, producing a consistent global reconstruction. During this process, the evolving geometric evidence updates the local instance libraries and progressively organizes them into a unified global 3D instance library. Experiments on long-sequence benchmarks show that our method produces more accurate trajectories and higher-quality 3D reconstructions, highlighting the effectiveness of persistent instance anchors for organizing long-horizon 3D reconstruction. Our code is available on the project page: https://yixn965.github.io/LIST3R/.

Problem

• long-sequence 3D reconstruction
• instance-aware
• fragment alignment
• global consistency
• spatial memory

Innovation

• instance-aware
• long-sequence 3D reconstruction
• instance anchors
• fragment alignment
• global 3D instance library