DLC-Bench: Detailed Localized Captioning Benchmark for Images and Videos

DLC-Bench is a dataset for benchmarking detailed and localized image and video captioning. It was created by researchers from NVIDIA, UC Berkeley, and UCSF, including Long Lian, Yifan Ding, and others. The dataset was last updated on the Hugging Face platform on April 24, 2025.

Use Cases

• Benchmarking detailed image captioning models based on localized description tasks.
• Evaluating vision-language model performance on generating spatially-aware captions.
• Training models for fine-grained scene understanding from visual data.

Strengths

• Created by a collaboration of major institutions including NVIDIA, UC Berkeley, and UCSF.
• Associated with a published paper, code, project page, and demo, indicating research rigor.
• Last updated on 2025-04-24, suggesting recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and dataset size are unknown, which may limit suitability assessment.
• The license is listed as CC BY-NC-SA 4.0, which restricts commercial use.

Provenance

Source

NVIDIA, UC Berkeley, UCSF

Collection Method

Likely created for research benchmarking as described in the associated paper.

Time Range

null

Freshness

Last updated 2025-04-24 17:43:52.

Geography

null