2026-08-12 · ECCV 2026

Can Vision-Language Models Assess Proxemic Risk from Egocentric Robot Images?

Vladyslava Rudas, Dmytro Kuzmenko

Published at ECCV 2026. 0 citations, as of the last refresh.

Abstract

Assessing proxemic danger from a robot's egocentric perspective is critical for safe embodied navigation in human environments and requires both visual and contextual reasoning. We evaluate three opensource vision-language models (VLMs) (\textit{InternVL}, \textit{Qwen-VL}, and \textit{SmolVLM}) on the classification of egocentric robot images into four danger levels, comparing three prompting strategies and two rounds of QLoRA fine-tuning against a stratified random baseline. Without fine-tuning, all models perform near the baseline, while fine-tuning yields only modest overall improvements. However, \textit{Qwen-VL} with an advanced prompt achieves substantially higher recall for high-danger cases than the other models. An analysis of person localization further shows that correct danger classification does not correspond to better spatial grounding, indicating that a model may produce a useful safety label without attending to the relevant region of the scene. These results show that current VLMs remain limited in fine-grained proxemic reasoning and spatial grounding, although targeted prompting and fine-tuning can improve high-danger detection in selected models.

arXiv comment: Accepted at the EMR 2026 workshop at ECCV 2026 (non-archival)

Ten-minute slide kit

Six slides is the whole talk: what was broken, what people tried, what these authors did, what the numbers say, where it falls over, and the sentence people should remember.

SLIDE 1

The problem

Mobile robots operating in environments shared with humans must constantly assess proxemic risk to navigate safely and avoid causing discomfort or collisions. A robot that is unable to correctly assess a crowded area, specifically, a high level of danger, or cannot distinguish it from an empty corridor, may invade personal space, fail to give way, or perform dangerous manoeuvres. Most existing approaches to safety assessment are based on explicit geometric reasoning or <span id="page-1-0"></span>specialised depth…

SLIDE 2

What came before

Not recoverable from the parsed text. Read this section in the paper yourself.

SLIDE 3

The method

SLIDE 4

What they measured

We report the following metrics for each model × fine-tuning stage × prompt type configuration: Accuracy – the fraction of correctly classified images. Weighted F1 – the average F1 score across all classes, weighted by the frequency of occurrence, which reflects the overall quality of the classification under class imbalance. Recall of high danger class – recall specifically for the high danger class, defined as the proportion of correctly identified scenes with high danger for the robot. This is a key safety…

SLIDE 5

Where it breaks

Not recoverable from the parsed text. Read this section in the paper yourself.

SLIDE 6

One-line takeaway

This work evaluates three opensource vision-language models on the classification of egocentric robot images into four danger levels, comparing three prompting strategies and two rounds of QLoRA fine-tuning against a stratified random baseline.

Assembled from the paper's own PDF, parsed with its layout intact so tables and equations survive, 27,976 characters of it, then split on the paper's own section headings. Extractive, not generated: every sentence here is lifted from the paper. Check it before you present it.