COCO-2017-Vietnamese: Translated Image Captions for Object Detection

A Vietnamese-language version of the COCO 2017 dataset, likely containing images with object annotations and translated captions. The dataset is maintained by AI Enthusiasm and was last updated on Hugging Face in January 2026. The original COCO dataset is a large-scale resource for object detection, segmentation, and captioning.

Use Cases

• Training Vietnamese image captioning models based on the translated text descriptions.
• Fine-tuning object detection models for Vietnamese contexts using the annotated images.
• Conducting multilingual computer vision research by comparing performance across language versions of COCO.

Strengths

• Based on the widely-used and benchmarked COCO 2017 dataset.
• Last updated on Hugging Face on 2026-01-11, suggesting recent maintenance.
• Provides a specific language adaptation (Vietnamese) for a major computer vision resource.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.
• The translation quality and coverage of the Vietnamese captions are not described.

Provenance

Source

AI Enthusiasm community, derived from the original COCO 2017 dataset.

Collection Method

Likely involves translation and adaptation of the original annotations.

Time Range

Based on the COCO 2017 version, the underlying data is from around 2017.

Freshness

Last updated 2026-01-11 08:06:55; freshness should be verified.

Geography

null