Spotlight Others

MooG: Self-supervised video model learns off-the-grid representations, enabling consistent scene element tracking even with motion; outperforming grid-based baselines on various vision tasks.

MotionBooth: A new framework enabling precise control over both object and camera movements in customized text-to-video generation, achieving high-quality video while maintaining training efficiency.

RealMotion: a novel motion forecasting framework for continuous driving that outperforms existing methods by accumulating historical scene information and sequentially refining predictions, achieving …

MCTS-transfer: Iteratively refining Bayesian optimization via Monte Carlo tree search for efficient black-box optimization using transfer learning.

Latent Prompt Transformer (LPT) revolutionizes molecule design by unifying generation and optimization, achieving high efficiency in discovering novel molecules with desired properties.

3D Gaussian Mapping (3DGM) achieves self-supervised camera-only 3D scene decomposition by leveraging multi-traverse driving data, memorizing permanent structures while filtering out transient objects.

MECD: A new task and dataset unlocks multi-event causal discovery in videos, enabling a novel framework that outperforms existing models by efficiently identifying causal relationships between chronol…

MambaTree: A novel tree-topology-based state space model surpasses existing methods by dynamically generating input-aware topologies for enhanced long-range dependencies in vision and language.

LLM-ESR enhances sequential recommendation by integrating semantic information from LLMs, significantly improving performance on long-tail users and items.

New algorithm efficiently solves linear regression with heterogeneous data batches, handling diverse input distributions and achieving high accuracy with fewer samples.

LightGaussian achieves 15x compression of 3D Gaussian scene representations, boosting rendering speed to 200+ FPS while maintaining visual quality, solving storage and efficiency issues in real-time n…

This paper introduces a novel training policy that leverages catastrophic forgetting to make diffusion models resilient against malicious fine-tuning, effectively preventing the generation of harmful …

This paper introduces a novel unsupervised video object segmentation method using long-term point trajectories and optical flow, outperforming prior art by effectively combining sparse, long-term moti…

Proper learning of noisy halfspaces with margins is achievable with sample complexity matching random classification noise, defying prior expectations.

A novel data-driven relighting model achieves state-of-the-art accuracy by learning latent intrinsic and extrinsic scene properties, even recovering albedo without explicit supervision.

LDNS: a new generative model for neural spiking data, enabling high-fidelity sampling and low-dimensional latent inference, paving the way for simulating realistic brain activity.

This paper proves that language generation in the limit is always possible, even with an adversarial setting, contrasting with the impossibility of language identification in the limit.

Kermut: A novel Gaussian process regression model achieves state-of-the-art accuracy in predicting protein variant effects and provides reliable uncertainty estimates, crucial for protein engineering …

Place3D optimizes LiDAR placement for superior 3D scene understanding.

This paper introduces L-Reg, a novel logical regularization technique, to improve generalization in visual classification. L-Reg effectively reduces model complexity and improves interpretability by f…