Picture for Jinwoo Kim

Jinwoo Kim

Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity

May 07, 2023
Jinghao Xin, Jinwoo Kim, Zhi Li, Ning Li

Figure 1 for Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity

Figure 2 for Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity

Figure 3 for Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity

Figure 4 for Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity

* 12 pages, 11 figures, 2 tables

Access Paper or Ask Questions

Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost

Oct 27, 2022
Sungjun Cho, Seonwoo Min, Jinwoo Kim, Moontae Lee, Honglak Lee, Seunghoon Hong

Figure 1 for Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost

Figure 2 for Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost

Figure 3 for Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost

Figure 4 for Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost

* 19 pages, 8 figures

Access Paper or Ask Questions

Equivariant Hypergraph Neural Networks

Aug 22, 2022
Jinwoo Kim, Saeyoon Oh, Sungjun Cho, Seunghoon Hong

Figure 1 for Equivariant Hypergraph Neural Networks

Figure 2 for Equivariant Hypergraph Neural Networks

Figure 3 for Equivariant Hypergraph Neural Networks

Figure 4 for Equivariant Hypergraph Neural Networks

* 29 pages, 2 figures

Access Paper or Ask Questions

Pure Transformers are Powerful Graph Learners

Jul 06, 2022
Jinwoo Kim, Tien Dat Nguyen, Seonwoo Min, Sungjun Cho, Moontae Lee, Honglak Lee, Seunghoon Hong

Figure 1 for Pure Transformers are Powerful Graph Learners

Figure 2 for Pure Transformers are Powerful Graph Learners

Figure 3 for Pure Transformers are Powerful Graph Learners

Figure 4 for Pure Transformers are Powerful Graph Learners

* 26 pages, 8 figures

Access Paper or Ask Questions

GraphDistNet: A Graph-based Collision-distance Estimator for Gradient-based Trajectory

Jun 03, 2022
Yeseung Kim, Jinwoo Kim, Daehyung Park

Figure 1 for GraphDistNet: A Graph-based Collision-distance Estimator for Gradient-based Trajectory

Figure 2 for GraphDistNet: A Graph-based Collision-distance Estimator for Gradient-based Trajectory

Figure 3 for GraphDistNet: A Graph-based Collision-distance Estimator for Gradient-based Trajectory

Figure 4 for GraphDistNet: A Graph-based Collision-distance Estimator for Gradient-based Trajectory

* 8 pages, 7 figures, submitted to RA-L with IROS 2022 Option

Access Paper or Ask Questions

Transformers Generalize DeepSets and Can be Extended to Graphs and Hypergraphs

Oct 27, 2021
Jinwoo Kim, Saeyoon Oh, Seunghoon Hong

Figure 1 for Transformers Generalize DeepSets and Can be Extended to Graphs and Hypergraphs

Figure 2 for Transformers Generalize DeepSets and Can be Extended to Graphs and Hypergraphs

Figure 3 for Transformers Generalize DeepSets and Can be Extended to Graphs and Hypergraphs

Figure 4 for Transformers Generalize DeepSets and Can be Extended to Graphs and Hypergraphs

* 21 pages, 5 figures

Access Paper or Ask Questions