Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Department of Digital Systems, University of Piraeus
2
0
3
Unlock better affect prediction: Language models can condition interpretable, handcrafted features to outperform end-to-end deep learning, offering a transparent and efficient alternative.
Stop throwing away valuable disagreement data: modeling annotation distributions with Beta distributions unlocks richer insights into emotional perception.