Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note3 min read

UESF-Bench: Benchmarking and Probing for Unified Embodied Seeking and Following

Source

arXiv

arxiv.org/abs/2607.13621v1

Read the source

Source

Plain-Language Brief

Researchers have introduced UESF-Bench, a new benchmark to evaluate the performance of AI agents in language-guided human following tasks. The current benchmarks assume that the target person is visible at the start of an episode, but this work focuses on scenarios where the agent needs to first find a language-described target and then persistently follow that target in a dynamic environment. The proposed framework, SeekFollow-VLA, uses a vision-language-action framework with a task-driven routing mechanism for latent phase inference and transition modeling between seeking and following.

What the Document Says

  • The paper introduces the Unified Embodied Seeking and Following Benchmark (UESF-Bench) to evaluate AI agents' performance in language-guided human following tasks.
  • The current benchmarks assume that the target person is visible at the start of an episode, while UESF-Bench focuses on scenarios where the agent needs to find a language-described target and then follow it persistently in a dynamic environment.
  • SeekFollow-VLA is a proposed framework using vision-language-action with a task-driven routing mechanism for latent phase inference and transition modeling between seeking and following.
  • The experimental results show that SeekFollow-VLA outperforms single-head and dual-head baselines across single-person and multi-person environments.

Key Concepts Mentioned

  • UESF-Bench
  • language-guided human following
  • AI agents
  • dynamic environment
  • SeekFollow-VLA
  • vision-language-action framework
  • latent phase inference
  • transition modeling

Claims Made by the Document

  • The current benchmarks oversimplify the language-guided human following problem.
  • UESF-Bench addresses this limitation by requiring agents to handle semantic-guided exploration, reliable behavior switching and recovery, and delayed identity grounding.
  • SeekFollow-VLA achieves clear improvements over single-head and dual-head baselines across single-person and multi-person environments.

Evidence Used

  • The text describes the limitations of current benchmarks and introduces UESF-Bench as a solution.
  • The text details the requirements of UESF-Bench and the features of SeekFollow-VLA.
  • The experimental results showing improvements of SeekFollow-VLA are presented.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Language-guided human following is an important capability for embodied agents, but existing benchmarks typically assume that the target person is visible at the start of an episode. This setting simplifies the problem and overlooks a more realistic requirement: an agent often needs to first find a language-described target and then persistently follow that target in a dynamic environment. While recent work has started to study human search, existing settings are typically evaluated in task-specific scenarios and often rely on stronger prior knowledge of the environment. Moreover, they usually treat searching and following as separate tasks and still lack a unified benchmark for systematic evaluation. To address these limitations, we introduce the Unified Embodied Seeking and Following Benchmark (UESF-Bench), a large-scale and diverse benchmark for embodied human seeking and following. The benchmark requires agents to handle semantic-guided exploration, reliable behavior switching and recovery, and delayed identity grounding. To this end, we propose SeekFollow-VLA, a vision-language-action framework with a task-driven routing mechanism for latent phase inference and transition modeling between seeking and following. Experimental results show that SeekFollow-VLA achieves clear improvements over both single-head and dual-head baselines across single-person and multi-person environments, establishing a baseline for unified embodied seek-and-follow.