Sign in to view source links and access this dataset
Description
WebEyes is a task-level benchmark for evaluating search-based visual reasoning, released by yangbokang81 and last updated on May 13, 2026. It supports three distinct datasets: WebEyes-Ground, WebEyes-Seg, and WebEyes-VQA. Each task is released as a JSONL file, with mirrored Parquet files used for direct image rendering on the Hugging Face platform.
Use Cases
Benchmarking visual grounding models based on the WebEyes-Ground dataset
Evaluating image segmentation models based on the WebEyes-Seg dataset
Testing visual question answering (VQA) systems based on the WebEyes-VQA dataset
Training multimodal AI agents for search-based reasoning tasks
Strengths
Provides a structured benchmark with three distinct, task-specific datasets
Includes mirrored Parquet files for direct image rendering on the Hugging Face platform
Released by a specific author (yangbokang81) with a paper and website referenced
Limitations
Column-level documentation is absent; field semantics must be inferred after download
Row count is unknown, which may limit suitability assessment
Data may reflect biases inherent to the source collection methods
Provenance
Source
yangbokang81
Freshness
Last updated 2026-05-13 07:37:43
License is unknown; terms of use must be verified before application.