Coco Synth Qwen3 Next 0 1: Synthetic COCO-Style Images with Scene Descriptions

coco_synth_qwen3_next_0_1 is a synthetic image dataset containing 10010 images. The dataset was created by author testdummyvt using an automated pipeline to mirror the COCO80 object list, pairing each generated image with a short scene description and structured object entries. It was last updated on Hugging Face on 2025-09-14.

Use Cases

• Train object detection models based on the COCO80 object list.
• Benchmark scene understanding algorithms based on the paired image and one-line scene descriptions.
• Develop or test multimodal (image-text) models based on the structured object entries and phrasings.
• Generate synthetic training data for computer vision pipelines based on the automated synthesis method.

Strengths

• Contains 10010 images, providing a substantial synthetic collection.
• Mirrors the established COCO80 object list, suggesting a structured benchmark.
• Pairs images with both scene-level and object-level text annotations.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count beyond the image total is unknown, which may limit suitability assessment.
• The synthetic nature may introduce biases not present in real-world data.

Provenance

Source

testdummyvt on Hugging Face.

Collection Method

Created with an automated pipeline that synthesizes object lists and one-line descriptions.

Time Range

null

Freshness

Last updated 2025-09-14 20:18:49.

Geography

null