小豪's Studio

ColorDynamic: Generalizable, Scalable, Real-time, End-to-end Local Planner for Unstructured & Dynamic Environments

ColorDynamic: 面向动态、非机构化环境的基于DRL的端到端局部规划算法

分享一下我的近期工作ColorDynamic，它由我之前的工作Color拓展而来，最核心的贡献是给模型赋予了时序感知能力，使得模型能够在非结构化的动态场景中实现局部规划。同时，该算法还可以无缝迁移到多机器人任务中。项目链接： https://github.com/XinJingHao/ColorDynamic 作者： Jinghao Xin, Zhichao Liang, Zih...

Posted by Jinghao on February 28, 2025

OkayPlan: A real-time path planning algorithm for dynamic environments

考虑障碍物运动学的动态实时路径规划

OkayPlan: Obstacle Kinematics Augmented Dynamic real-time path Planning via particle swarm optimization 期刊： Ocean Engineering [Q1, TOP, IF:4.6] 作者： Xin J, Kim J, Chu S, et al...

Posted by Jinghao on April 7, 2024

Color: Train a Real-world Local Path Planner in One Hour

DRL与AGV能碰撞出怎样的火花?

Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and ectorized Diversity 期刊： Engineering Applications of Artificial Intelligence 作...

Posted by Jinghao on January 17, 2024

SEPSO: Self-Evolving Particle Swarm Optimization for Real-time Path Planning

基于PSO的实时路径规划

Efficient Real-Time Path Planning with Self-Evolving Particle Swarm Optimization in Dynamic Scenarios 期刊： Unmanned Systems [Q2, IF:3.0] 作者： Xin J, Li Z, Zhang Y, et al. ...

Posted by Jinghao on December 22, 2023

向量的叉乘、点乘，看这一篇就够了！

知乎高质量文章

摘自我的知乎文章向量的叉乘、点乘，看这一篇就够了！

Posted by Jinghao on December 18, 2023

这一刻，我终于站在了孤独星球的封面里

我的日常

这里是印度尼西亚的Bromo火山，被评为世界上最像月球表面的地方瞌睡是晚上12点睡的，火山是凌晨3点起来爬的，只为看「那一抹日出」…

Posted by Jinghao on August 26, 2023

「付代码」Prioritized Experience Replay(优先经验回放)和Sum-tree解析

知乎高质量文章

摘自我的知乎文章「付代码」Prioritized Experience Replay(优先经验回放)和Sum-tree解析

Posted by Jinghao on May 22, 2023

Hello, Singapore!

我的日常

忙了大半年的访学流程，进度条终于达到了100% 接下来就要换个地方搬砖了…

Posted by Jinghao on May 1, 2023

「付代码」Soft Actor Critic 详细推导与深入理解

知乎高质量文章

摘自我的知乎文章「付代码」Soft Actor Critic 详细推导与深入理解

Posted by Jinghao on September 23, 2022

「强化学习」PPO的理论推导

知乎高质量文章

摘自我的知乎文章「强化学习」PPO的理论推导

Posted by Jinghao on September 14, 2022

Hi! It's Jinghao

ColorDynamic: Generalizable, Scalable, Real-time, End-to-end Local Planner for Unstructured & Dynamic Environments

ColorDynamic: 面向动态、非机构化环境的基于DRL的端到端局部规划算法

OkayPlan: A real-time path planning algorithm for dynamic environments

考虑障碍物运动学的动态实时路径规划

Color: Train a Real-world Local Path Planner in One Hour

DRL与AGV能碰撞出怎样的火花?

SEPSO: Self-Evolving Particle Swarm Optimization for Real-time Path Planning

基于PSO的实时路径规划

向量的叉乘、点乘，看这一篇就够了！

知乎高质量文章

这一刻，我终于站在了孤独星球的封面里

我的日常

「付代码」Prioritized Experience Replay(优先经验回放)和Sum-tree解析

知乎高质量文章

Hello, Singapore!

我的日常

「付代码」Soft Actor Critic 详细推导与深入理解

知乎高质量文章

「强化学习」PPO的理论推导

知乎高质量文章

FEATURED TAGS

ABOUT ME

FRIENDS