Aligned-8-Emotion: Multimodal Dataset with English and Amharic Text and 88,360 Face Images

Aligned-8-Emotion-Dataset-Final is a multimodal dataset containing 88,360 face images and text in both English and Amharic, annotated for 8 emotion categories. The dataset appears to be sourced from Kaggle, but specific authorship, collection methodology, and temporal details are not provided. Its primary purpose is likely for training and evaluating emotion recognition models across different data modalities and languages.

Use Cases

• Train facial emotion classifiers based on the 88,360 annotated face images.
• Develop multimodal emotion recognition models based on aligned text and image data.
• Build or benchmark cross-lingual NLP models for emotion detection based on the English and Amharic text data.
• Conduct research on the alignment between textual descriptions and visual expressions of emotion.

Strengths

• Dataset contains 88,360 face images, providing a substantial visual corpus.
• Data is multimodal, combining image and text modalities for aligned analysis.
• Supports two languages (English and Amharic), enabling cross-lingual research.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the text component is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality and annotation consistency require manual inspection.