EffiNav: Fusing Depth and Vision-Language for Efficient Object Goal Navigation
Overview
This paper introduces EffiNav, a method that combines depth sensing with vision-language models to achieve efficient object goal navigation in unknown environments. The authors evaluate the approach on two simulation benchmarks—Habitat Matterport 3D (HM3D) and Open-Vocabulary Object goal Navigation (OVON)—and further validate its performance on physical robots in real-world settings. Failure analysis is conducted across large-scale simulation episodes, and the method is extended to a memory-augmented navigation task on the GOAT-BENCH dataset. Results show that EffiNav matches or outperforms recent baselines on standard metrics including Success Rate (SR) and Success weighted by Path Length (SPL), demonstrating its efficiency, robustness, and generalizability for autonomous exploration tasks.
Generated from attributed reports · Updated 22 minutes ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
No current evidence-backed claims. Missing information remains not reported.
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalEffiNav: Fusing Depth and Vision-Language for Efficient Object Goal Navigation
This paper introduces EffiNav, a method that combines depth sensing with vision-language models to achieve efficient object goal navigation in unknown environments. The authors evaluate the approach on two simulation benchmarks—Habitat Matterport 3D (HM3D) and Open-Vocabulary Object goal Navigation (OVON)—and further validate its performance on physical robots in real-world settings. Failure analysis is conducted across large-scale simulation episodes, and the method is extended to a memory-augmented navigation task on the GOAT-BENCH dataset. Results show that EffiNav matches or outperforms recent baselines on standard metrics including Success Rate (SR) and Success weighted by Path Length (SPL), demonstrating its efficiency, robustness, and generalizability for autonomous exploration tasks.
Event coverage history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).