Search papers, labs, and topics across Lattice.
Department of Computer Science, American International University–Bangladesh
2
0
5
Stop relying solely on text or UI features for app rating prediction: a lightweight vision-language framework achieves state-of-the-art results by fusing MobileNetV3 visual features with DistilBERT textual features.
Achieve 2-15% better text classification accuracy by explicitly fusing grammatical features with frozen contextual embeddings, offering a lightweight alternative to computationally intensive transformer models.
