AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

360CityArena: A Realistic Virtual Urban Navigation Benchmark for Embodied Agents

Kenta Watanabe, Atsuyuki Miyai, Mizuki Takenawa, Kiyoharu Aizawa, Toshihiko Yamasaki

ECCV 2026 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Vision-language navigation agents · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
add location information (map marker) to observation (ablation)GPT-5 · 360CityArena Landmark (Lang)success rate (%): 24.0 -> 16.0Hurt
add location information (map marker) to observation (ablation)GPT-5 · 360CityArena Rel Reasonsuccess rate (%): 48.0 -> 32.0Hurt
add location information (map marker) to observation (ablation)GPT-5 · 360CityArena Landmark (Img)success rate (%): 44.0 -> 48.0No clear effect
add location information (map marker) to observation (ablation)GPT-5 · 360CityArena Map Navsuccess rate (%): 4.0 -> 0.0No clear effect
add location information (map marker) to observation (ablation)GPT-5 · 360CityArena VLNsuccess rate (%): 8.0 -> 8.0No clear effect
add location information (map marker) to observation (ablation)GPT-5 · 360CityArena Obj Countsuccess rate (%): 0.0 -> 2.4Improved
use landmark image instead of language descriptionGPT-5 · 360CityArena Landmark Searchsuccess rate (%): 16.0 -> 48.0Improved
use landmark image instead of language descriptionClaude Sonnet 4.5 · 360CityArena Landmark Searchsuccess rate (%): 4.0 -> 16.0Improved
use landmark image instead of language descriptionQwen2.5-VL-32B-Instruct · 360CityArena Landmark Searchsuccess rate (%): 16.0 -> 20.0Improved
use landmark image instead of language descriptionInternVL3.5-8B · 360CityArena Landmark Searchsuccess rate (%): 20.0 -> 20.0No clear effect
use landmark image instead of language descriptionInternVL3.5-38B · 360CityArena Landmark Searchsuccess rate (%): 0.0 -> 16.0Improved
increase task difficulty from Easy to Medium to Hard (Environment Understanding)GPT-5 · 360CityArena Environment Understandingsuccess rate (%): 28.0 -> 25.0 -> 18.5Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy