AMC Attention Guided Multi-modal Image Search Dataset

AMC is a dataset for attention-guided multi-modal correlation learning in image search. The dataset was created by kanchen-usc and last updated in July 2019. Specific details on volume, format, and content are not provided.

Use Cases

• Train attention models to correlate image and text features for search.
• Benchmark multi-modal learning algorithms for retrieval tasks.
• Study the effectiveness of attention mechanisms in aligning visual and textual data.

Strengths

• Released under the permissive MIT license.
• Focuses on a specific research area: attention-guided multi-modal correlation.

Limitations

• No information on dataset size, rows, or columns.
• No sample data or file formats are described.
• Data is from 2019 and may not reflect current benchmarks.

Provenance

Source

github repository by kanchen-usc

Collection Method

null

Time Range

null

Freshness

Last updated in 2019.

Geography

null