ShuffleNet-V2: Network Structure for Miner Emotion Recognition

A dataset containing five categories of miners' unsafe emotions used to train a bimodal deep learning model. The model, developed by Ying Lu, achieved a mean recognition accuracy of 85.56% and was preliminarily tested in a real mining environment. The dataset was last updated on 2026-05-15.

Use Cases

• Training emotion recognition models based on labeled categories of unsafe emotions.
• Developing multimodal frameworks based on integrated speech and facial expression features.
• Benchmarking deep learning models like ShuffleNet-V2 for feature extraction tasks.
• Conducting preliminary field tests of AI models in real-world industrial conditions.

Strengths

• Model achieved a mean recognition accuracy of 85.56% on the dataset.
• Dataset supports a bimodal framework integrating both speech and facial expression features.
• Includes results from a preliminary field test in a real mining environment.

Limitations

• Dataset size is very small at 5.5 KB, indicating limited scope.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Ying Lu via figshare.

Collection Method

Constructed for network training of a deep learning model.

Freshness

Last updated 2026-05-15 17:37:00; freshness should be verified.