EAC-Agent Ablation Study Results on IEMOCAP and MELD

Ablation study results for the EAC-Agent multimodal conversational model, published by Shahid Jamil on figshare in April 2026. The dataset likely contains performance metrics from experiments comparing the model's emotion recognition and response generation capabilities against benchmarks. The proposed model achieved an accuracy of 76.27% on IEMOCAP and 67.57% on MELD for emotion recognition.

Use Cases

• Benchmarking multimodal emotion recognition models based on reported accuracy scores
• Evaluating emotion-aware response generation based on reported perplexity, BLEU, and ROUGE-L scores
• Comparing transformer-based sequence-to-sequence architectures for conversational agents
• Studying the integration of text, audio, and visual features in AI models

Strengths

• Model performance metrics are provided with specific numbers (e.g., 76.27% accuracy)
• Results are validated on two benchmark datasets (IEMOCAP and MELD)
• The dataset is openly licensed under CC-BY-4.0

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• The dataset is only 5.5 KB, indicating a very limited scope likely containing only summary results

Provenance

Source

figshare

Collection Method

Likely contains experimental results from an ablation study of the EAC-Agent model.

Freshness

Last updated 2026-04-17 17:35:28