LHPR-VLN: Long-Horizon Vision-Language Navigation Benchmark

LHPR-VLN is a benchmark for long-horizon planning and reasoning in Vision-Language Navigation, introduced in the paper 'Towards Long-Horizon Vision-Language Navigation: Platform, Benchmark and Method'. The dataset is hosted by author Starry123 on Hugging Face and was last updated on May 30, 2025. It is also available on the ModelScope platform.

Use Cases

• Training agents for long-horizon navigation based on the described trajectory and step-task data.
• Benchmarking planning and reasoning algorithms for vision-language navigation tasks.
• Developing methods to decompose complex navigation instructions into sequential actions.
• Studying agent failure modes in extended, multi-step environments.

Strengths

• Dataset is explicitly designed for the challenging domain of long-horizon vision-language navigation.
• Associated with a published research paper, providing methodological context.
• Last updated on 2025-05-30, indicating recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and sample data are unknown, which may limit suitability assessment.
• License information is unavailable, which could restrict usage.

Provenance

Source

Starry123

Collection Method

Introduced as a benchmark in the associated research paper; likely contains simulated or recorded navigation trajectories.

Time Range

null

Freshness

Last updated 2025-05-30 12:29:06.

Geography

null