10 accepted papers
Mixture-of-Experts (MoE) models scale by activating only a small subset of experts per token, but standard deterministic top-$k$ routing is non-differentiable and trained using surrogate gradients that ignore the discrete expert selection used at inference. We introduce SIMoE routing by modeling exp…
Probing learned concepts in large language models (LLMs) is crucial for understanding how semantic knowledge is encoded internally. Training linear classifiers on probing tasks is a principle approach to denote the vector of a certain concept in the representation space. However, the single vector i…
Implicit Neural Representations (INRs) are continuous function learners for conventional digital signal representations. With the aid of positional embeddings and/or exhaustively fine-tuned activation functions, INRs have surpassed many limitations of traditional discrete representations. However, e…
This paper presents a universal LiDAR point cloud global localization framework based on multi-sector overlapping loss to address the localization challenges caused by heterogeneous LiDAR point clouds with varying resolutions, scanning formats, and field of view differences. The proposed method firs
One-shot global localization is crucial in many robotic applications, providing significant advantages during initialization and relocalization processes. However, LiDAR-based one-shot global localization methods encounter challenges, including local feature matching errors, sensitivity to dynamic o
Implicit Neural Representations (INRs) provide a continuous mapping between the coordinates of a signal and the corresponding values. As the performance of INRs heavily depends on the choice of nonlinear-activation functions, there has been a significant focus on encoding explicit signals within INR…
This paper presents a novel one-shot global localization algorithm based on semantic-guided triangle descriptors to address initialization and global localization challenges in GNSSdenied environments. By encoding semantic geometric information into triangle descriptors, the proposed approach achiev
Large language models (LLMs) have demonstrated exceptional capabilities across a wide range of tasks, from text generation to complex problem-solving. LLM APIs provide easy access to these models by streamlining deployment and usage. Combining LLMs with complementary strengths has been shown to yiel…
Weakly-supervised Spatio-Temporal Video Grounding(STVG) aims to localize target object tube given a text query, without densely annotated training data. Existing methods extract each candidate tube feature independently by cropping objects from video frame feature, discarding all contextual informat…
Current 6D pose estimation methods focus on handling objects that are previously trained, which limits their applications in real dynamic world. To this end, we propose a geometry correspondence-based framework, termed GCPose, to estimate 6D pose of arbitrary unseen objects without any re-training.…