Search papers, labs, and topics across Lattice.
6
29
8
11
Calibration accuracy improves dramatically when using PLS regression, solving issues that plague traditional methods in constrained robotic environments.
HSAN redefines VLN-CE by combining advanced semantic scene representation with optimal transport planning, achieving unprecedented navigation success in complex environments.
MMRAG gets a human-like reasoning upgrade: CogniVerse uses cognitive reflection and information geometry to filter noise, align modalities, and generate coherent responses, outperforming existing systems.
LVLMs are surprisingly susceptible to universal, black-box adversarial attacks that synergistically combine imperceptible image perturbations with subtle text prompts.
VLMs can be significantly improved by reasoning over diverse, generated text inputs, rather than relying on restrictive, predefined templates.
VLMs can now handle real-world sensor failures and data privacy constraints without catastrophic performance drops, thanks to a new plug-and-play module for incomplete multi-modal inputs.
