paper-with-me

홈 › Papers

AMEND: A Mixture of Experts Framework for Long-tailed Trajectory Prediction

2024-02-13 · Ray Coden Mercurius, Ehsan Ahmadi, Soheil Mohamad Alizadeh Shabestary, Amir Rasouli

Accurate prediction of pedestrians' future motions is critical for intelligent driving systems. Developing models for this task requires rich datasets containing diverse sets of samples. However, the existing naturalistic trajectory prediction datasets are generally imbalanced in favor of simpler samples and lack challenging scenarios. Such a long-tail effect causes prediction models to underperform on the tail portion of the data distribution containing safety-critical scenarios. Previous methods tackle the long-tail problem using methods such as contrastive learning and class-conditioned hypernetworks. These approaches, however, are not modular and cannot be applied to many machine learning architectures. In this work, we propose a modular model-agnostic framework for trajectory prediction that leverages a specialized mixture of experts. In our approach, each expert is trained with a specialized skill with respect to a particular part of the data. To produce predictions, we utilise a router network that selects the best expert by generating relative confidence scores. We conduct experimentation on common pedestrian trajectory prediction datasets and show that our method improves performance on long-tail scenarios. We further conduct ablation studies to highlight the contribution of different proposed components.

📄 PDF Abstract BibTeX arXiv:2402.08698

Code (0)

등록된 구현이 없습니다.

Tasks

Contrastive LearningMixture-of-ExpertsPedestrian Trajectory PredictionPredictionTrajectory Prediction

Methods 이 논문이 사용한 방법론

Contrastive Learning 설명 없음

Similar Papers 제목 키워드 기반

LPT++: Efficient Training on Mixture of Long-tailed Experts

2024-09-17 · Bowen Dong, Pan Zhou, WangMeng Zuo

We introduce LPT++, a comprehensive framework for long-tailed classification that combines parameter-efficient fine-tuning (PEFT) with a learnable model ensemble. LPT++ enhances frozen Vision Transformers (ViTs) through …

Mixture-of-Expertsparameter-efficient fine-tuning

Preserving Long-Tailed Expert Information in Mixture-of-Experts Tuning

2026-04-24 · Haoze He, Xingyuan Ding, Xuan Jiang, Xinkai Zou 외 arxiv

Despite MoE models leading many benchmarks, supervised fine-tuning (SFT) for the MoE architectures remains difficult because its router layers are fragile. Methods such as DenseMixer and ESFT mitigate router collapse wit…

Mathematical Reasoning

Long-Tailed Distribution-Aware Router For Mixture-of-Experts in Large Vision-Language Model

2025-07-02 · Chaoxiang Cai, Longrong Yang, Minghe Weng, Xuewei Li 외 arxiv

The mixture-of-experts (MoE) architecture, which replaces dense networks with sparse ones, has attracted significant attention in large vision-language models (LVLMs) for achieving comparable performance while activating…

Long-Tailed Visual Recognition via Self-Heterogeneous Integration with Knowledge Excavation

2023-04-03 · CVPR 2023 1 · Yan Jin, Mengke Li, Yang Lu, Yiu-ming Cheung 외

Deep neural networks have made huge progress in the last few decades. However, as the real-world data often exhibits a long-tailed distribution, vanilla deep models tend to be heavily biased toward the majority classes. …

Mixture-of-ExpertsTransfer Learning

Toward Inference-optimal Mixture-of-Expert Large Language Models

2024-04-03 · Longfei Yun, Yonghao Zhuang, Yao Fu, Eric P Xing 외

Mixture-of-Expert (MoE) based large language models (LLMs), such as the recent Mixtral and DeepSeek-MoE, have shown great promise in scaling model size without suffering from the quadratic growth of training cost of dens…