自己教師あり物体ゴールナビゲーションと実環境でのオンライン微調整
Self-Supervised Object Goal Navigation with In-Situ Finetuning
位置一貫性を自己教師信号として使うことで、ラベル付き3Dメッシュに頼らず物体ナビゲーションエージェントを学習し、実環境でのオンライン微調整も可能にした研究。
著者: So Yeon Min, Yao-Hung Hubert Tsai, Wei Ding, Ali Farhadi, Ruslan Salakhutdinov, Yonatan Bisk, Jian Zhang
分類: cs.RO, cs.LG
原文アブストラクト
A household robot should be able to navigate to target objects without requiring users to first annotate everything in their home. Most current approaches to object navigation do not test on real robots and rely solely on reconstructed scans of houses and their expensively labeled semantic 3D meshes. In this work, our goal is to build an agent that builds self-supervised models of the world via exploration, the same as a child might - thus we (1) eschew the expense of labeled 3D mesh and (2) enable self-supervised in-situ finetuning in the real world. We identify a strong source of self-supervision (Location Consistency - LocCon) that can train all components of an ObjectNav agent, using unannotated simulated houses. Our key insight is that embodied agents can leverage location consistency as a self-supervision signal - collecting images from different views/angles and applying contrastive learning. We show that our agent can perform competitively in the real world and simulation. Our results also indicate that supervised training with 3D mesh annotations causes models to learn simulation artifacts, which are not transferrable to the real world. In contrast, our LocCon shows the most robust transfer in the real world among the set of models we compare to, and that the real-world performance of all models can be further improved with self-supervised LocCon in-situ training.
関連論文
- 動く前に探索方針を決める:拡散モデル生成の探索参照を用いた半静的環境でのゼロショット物体ナビゲーション物体ナビゲーション
- HTCE:同相写像に着想を得た物体ナビゲーションのためのシーン横断経験の正規化と再利用物体ナビゲーション
- SOR-Nav: 探索継続か領域移動か?文脈ゲート付き探索と領域間移動による物体ナビゲーション物体ナビゲーション
- NaviScale: 物体ナビゲーション向け大規模セマンティックマップデータセットの生成物体ナビゲーション
- ユーザー中心物体ナビゲーション:個人の習慣を統合したパーソナライズド物体探索ベンチマーク物体ナビゲーション
- Hydra-Nav: 適応的二重プロセス推論による物体ナビゲーション物体ナビゲーション
