Search papers, labs, and topics across Lattice.
7
0
7
4
AHRA reveals that aligning multimodal inputs at the same semantic depth can lead to significant performance drops, especially under noisy conditions.
MotionCraft achieves high-quality video super-resolution with predictable control over temporal smoothness and fidelity, outperforming traditional methods in complex motion scenarios.
Explicitly modeling modality reliability can drastically enhance sentiment analysis performance in the face of incomplete multimodal data.
Multimodal sentiment analysis suffers from "branch imbalance," where shared representations become redundant and private representations lose discriminative power, but a new rebalancing framework can fix it.
VLMs can be devastatingly fooled by modifying less than 2% of image pixels in a fixed, X-shaped pattern, causing them to fail spectacularly across diverse tasks like classification, captioning, and question answering.
By explicitly modeling a multi-level semantic hierarchy and carefully controlling information exchange between modalities, CLCR achieves state-of-the-art results in multimodal learning tasks ranging from emotion recognition to action recognition.
Unlocking superior multimodal sentiment analysis, TSD reveals that disentangling features into common, pairwise, and private subspaces dramatically boosts performance.
