Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics

  • 类型:arxiv
  • 标识:1705.07115
  • 链接:https://arxiv.org/abs/1705.07115
  • 主题:multimodal
  • 主分类:engineering
  • 形态:method
  • 被引:4348
  • 被引来源:Semantic Scholar
  • S2被引:4348
  • OpenAlex被引:500
  • 影响力被引:398
  • TLDR:A principled approach to multi-task deep learning is proposed which weighs multiple loss functions by considering the homoscedastic uncertainty of each task, allowing us to simultaneously learn various quantities with different units or scales in both classification and regression settings.
  • OpenAlex ID:W2618011341
  • OpenAlex DOI:10.48550/arxiv.1705.07115
  • DOI:10.48550/arxiv.1705.07115
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://arxiv.org/pdf/1705.07115
  • OpenAlex更新:2026-08-07
  • 待LLM分类:否
  • 成熟度:research
  • 场景:multi-task learning、loss weighting、uncertainty estimation
  • 标题中文:利用不确定性为损失加权的多任务学习,用于场景几何与语义
  • TLDR中文:本文提出一种多任务深度学习的原则性方法,通过考虑各任务的同方差不确定性来加权多个损失函数,从而在分类与回归场景下同时学习具有不同单位或尺度的多种量。
  • 来源文件
  • [OpenAlex discover]
  • [S2 enrich]