Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Kouichi SakuraiSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works77 from public data
- Personalize Segment Anything Model with One Shot400
A training-free Personalization approach for SAM, termed as PerSAM, which effectively adapt SAM for private use without any training, and can enhance DreamBooth to personalize Stable Diffusion for text-to-image generation, which discards the background disturbance for better target appearance learning.
- 383
- ManipLLM: Embodied Multimodal Large Language Model for Object-Centric Robotic Manipulation264
This work introduces an innovative approach for robot manipulation that leverages the robust reasoning capabilities of Multimodal Large Language Models (MLLMs) to enhance the stability and generalization of manipulation.
- SpatialBot: Precise Spatial Understanding with Vision Language Models220
SpatialBot, a model designed to enhance spatial understanding by utilizing both RGB and depth images, is proposed and the SpatialQA and SpatialQA $\boldsymbol{E}$ datasets are introduced to train VLMs for depth perception.
- Predictive Inverse Dynamics Models are Scalable Learners for Robotic Manipulation204
By closing the loop between vision and action, the end-to-end PIDM can be a better scalable action learner, and is initially pre-trained on large-scale robotic datasets, such as DROID, and can be adapted to realworld scenarios with a little fine-tuning data.
- Instruct2Act: Mapping Multi-modality Instructions to Robotic Actions with Large Language Model177
This paper presents Instruct2Act, a framework that utilizes Large Language Models to map multi-modal instructions to sequential actions for robotic manipulation tasks, employing the LLM model to generate Python programs that constitute a comprehensive perception, planning, and action loop for robotic tasks.
- Bridging Zero-shot Object Navigation and Foundation Models through Pixel-Guided Navigation Skill151
This work proposes a Pixel-guided Navigation skill (PixNav), which bridges the gap between the foundation models and the embodied navigation task, and designs an LLM-based planner to utilize the commonsense knowledge between objects and rooms to select the best waypoint.
- GFPose: Learning 3D Human Pose Prior with Gradient Fields102
GFPose is a versatile framework to model plausible 3D human poses for various applications, which is a time-dependent score network, which estimates the gradient on each body joint and progressively denoises the perturbed3D human pose to match a given task specification.
- Skill Reinforcement Learning and Planning for Open-World Long-Horizon Tasks92
This work proposes three types of basic skills in Minecraft, and uses RL with intrinsic rewards to accomplish basic skills with high success rates and outperforms baselines in most tasks by a large margin.
- SimMMDG: A Simple and Effective Framework for Multi-modal Domain Generalization80
This paper argues that mapping features from different modalities into the same embedding space impedes model generalization and proposes SimMMDG, a simple yet effective multi-modal DG framework that achieves strong performance on the EPIC-Kitchens dataset and the novel Human-Animal-Cartoon dataset.
- 75
- A Survey of Reasoning with Foundation Models73
This paper introduces seminal foundation models proposed or adaptable for reasoning, highlighting the latest advancements in various reasoning tasks, methods, and benchmarks and discusses the relevance of multimodal learning, autonomous agents, and super alignment in the context of reasoning.
- ManipVQA: Injecting Robotic Affordance and Physically Grounded Information into Multi-Modal Large Language Models57
ManipVQA is proposed, a novel framework that infuses MLLMs with manipulation-centric knowledge through a Visual Question-Answering (VQA) format and preserves the original vision-reasoning abilities while incorporating the newly acquired robotic insights.
- 1002 km twin-field quantum key distribution with finite-key analysis43
By optimizing parameters for shorter fiber distances, the feasibility of secure QKD over a 1002 km fibre channel considering the finite size effect is demonstrated and performance tests on key distribution for fiber lengths ranging from 202 km to 505 km are conducted.
- SparseDFF: Sparse-View Feature Distillation for One-Shot Dexterous Manipulation37
SarseDFF is presented, a novel DFF for 3D scenes utilizing large 2D vision models to extract semantic features from sparse RGBD images, a domain where research is limited despite its relevance to many tasks with fixed-camera setups.
- Heterogeneous-Agent Mirror Learning: A Continuum of Solutions to Cooperative MARL35
A novel framework named Heterogeneous-Agent Mirror Learning (HAML) is introduced that provides a general template for MARL algorithmic designs and it is proved that algorithms derived from the HAML template satisfy the desired properties of the monotonic improvement of the joint reward and the convergence to Nash equilibrium.
- 31
- 30
- Improving Compositional Text-to-image Generation with Large Vision-Language Models30
The proposed methodology significantly improves text-image alignment in compositional image generation, particularly with respect to object number, attribute binding, spatial relationships, and aesthetic quality.
- 27
- 27
- Towards Multimodal Open-Set Domain Generalization and Adaptation through Self-supervision25
This work introduces a novel approach to address Multimodal Open-Set Domain Generalization for the first time, utilizing self-supervision, and introduces two innovative multimodal self-supervised pretext tasks: Masked Cross-modal Translation and Multimodal Jigsaw Puzzles.
- 25
- A repeated game approach for analyzing the collusion on selective forwarding in multihop wireless networks22
This paper first puts forward a sub-route oriented punish and reward scheme, and proposes an multi-attacker repeated colluding game, and finds the sub-game equilibriums which indicate the attackers' optimal attack strategies.
- Hierarchical Interdisciplinary Topic Detection Model for Research Proposal Classification19
A deep Hierarchical Interdisciplinary Research Proposal Classification Network (HIRPCN) is developed, which proposes a hierarchical transformer to extract the textual semantic information of proposals and designs a level-wise prediction component to fuse the two types of knowledge representations and detect interdisciplinary topic paths for each proposal.
- Towards Robust Multimodal Open-set Test-time Adaptation via Adaptive Entropy-aware Optimization17
This work presents Adaptive Entropy-aware Optimization (AEO), a novel framework specifically designed to tackle Multimodal Open-set Test-time Adaptation (MM-OSTTA) for the first time, and shows that the entropy difference between known and unknown samples in the target domain strongly correlates with MM-OSTTA performance.
- 17
- Education robot object detection with a brain-inspired approach integrating Faster R-CNN, YOLOv3, and semi-supervised learning16
This paper introduces the progress of an education robot's object detection based on a brain-inspired heuristic method, which integrates Faster R-CNN, YOLOv3, and semi-supervised learning and shows that the proposed optimization algorithm significantly outperforms individual algorithms in terms of accuracy and real-time performance.
- 15
- Learning-Based Dimensionality Reduction for Computing Compact and Effective Local Feature Descriptors13
The experiments show that the lightweight MLPs trained using supervised method achieve better dimensionality reduction than PCA, and the lower-dimensional descriptors generated by the approach outperform the original higher- dimensional descriptors in downstream tasks, especially for the hand-crafted ones.
- 10
- Recall and Refine: A Simple but Effective Source-free Open-set Domain Adaptation Framework10
Recall and Refine (RRDA) is proposed, a novel SF-OSDA framework designed to address limitations by explicitly learning features for target-private unknown classes by addressing both domain shifts and distinguishability to unknown classes.
- MARLlib: A Scalable and Efficient Multi-agent Reinforcement Learning Library10
This paper proposes Multi-Agent RLlib (MARLlib), a comprehensive MARL algorithm library that facilitates RLlib for solving multi-agent problems and goes beyond current work by integrating diverse environment interfaces and providing flexible parameter sharing strategies.
- CrayonRobo: Object-Centric Prompt-Driven Vision-Language-Action Model for Robotic Manipulation9
This work introduces CrayonRobo, a model that leverages comprehensive multi-modal prompts that explicitly convey both low-level actions and high-level planning in a simple manner and enhances its robustness on unseen tasks by providing easily interpretable prompts.
- EqvAfford: SE(3) Equivariance for Point-Level Affordance Learning8
This work proposes the EqvAfford framework, with novel designs to guarantee the equivariance in point-level affordance learning for downstream robotic manipulation, with great performance and generalization ability on representative tasks on objects in diverse poses.
- 8
- Collaboration of Pre-trained Models Makes Better Few-shot Learner8
This paper proposes CoMo, a Collaboration of pre-trained Models that incorporates diverse prior knowledge from various pre-training paradigms for better few-shot learning and introduces a learnable Multi-Knowledge Adapter (MK-Adapter) to adaptively blend the predictions from CLIP and DINO.
- 7
- 6
- Mixed-Strategy Game Based Trust Management for Clustered Wireless Sensor Networks6
This paper proposes a novel, integrated trust management mechanism for the cluster wireless sensor networks, and analyzes the optimal decision making policy by using game theory to determine the optimal utility of the WSNs.
- Who Should Review Your Proposal? Interdisciplinary Topic Path Detection for Research Proposals5
A deep Hierarchical Interdisciplinary Research Proposal Classification Network (HIRPCN) is developed, which proposes a hierarchical transformer to extract the textual semantic information of proposals and designs a level-wise prediction component to fuse the two types of knowledge representations and detect interdisciplinary topic paths for each proposal.
- A Game Theory-Based Surveillance Mechanism against Suspicious Insiders in MANETs5
This work utilizes a reputation scheme to distinguish suspicious insiders from legitimate members, and quantify the threat level of the suspicious insiders, and utilizes game theory to analyze the interaction between monitor and suspicious insider.
- ManipGPT: Is Affordance Segmentation by Large Vision Models Enough for Articulated Object Manipulation?3
ManipGPT is introduced, a framework designed to predict optimal interaction areas for articulated objects using a large pretrained vision transformer (ViT), and a dataset of 9.9k simulated and real images is created to bridge the visual simto-real gap and enhance real-world applicability.
- A transit network design and frequency setting model with graph neural network and deep reinforcement learning3
This paper introduces a novel method for addressing the challenges associated with the Transit Network Design and Frequency Setting Problem (TNDFSP), which takes a simultaneous optimization approach leveraging graph neural networks and deep reinforcement learning to optimize passenger benefits, operating costs, and service frequencies.
- MCTS-EP: Empowering Embodied Planning with Online Preference Optimization2
It is theoretically prove that MCTS-EP achieves better performance bounds than conventional on-policy algorithms when the loss function is strongly convex, and it is demonstrated that it can be formulated as a search-enhanced variant of GAIL.
- SpikeStereoNet: A Brain-Inspired Framework for Stereo Depth Estimation from Spike Streams2
SpikeStereoNet is a brain-inspired framework and the first to estimate stereo depth directly from raw spike streams, fuses raw spike streams from two viewpoints and iteratively refines depth estimation through a recurrent spiking neural network (RSNN) update module.
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- SR3D: Unleashing Single-view 3D Reconstruction for Transparent and Specular Object Grasping1
This work proposes view matching and keypoint matching mechanisms, which leverage both the 2D and 3D’s inherent semantic and geometric information in the observation to determine the object’s 3D state within the scene, thereby reconstructing an accurate 3D depth map for effective grasp detection.
- 1
- 1
- Symmetric ternary quantum Fourier transform and its application1
The research of ternary quantum system has gradually come into the attention of scholars in recent years, and the quantum Fourier transform (QFT) offers an interesting way to perform arithmetic operations on a quantum computer.
- 1
- 1
- 1
- –
- –
- Research on Temporal Structures and Cultural Memory Mechanisms in Digital Exhibition Design of Intangible Cultural Heritage–
The study proposes a three-layer temporal structure model—authentic time, representational time, and experiential time—and summarizes the temporal translation methods found in various media forms, including video, immersive environments, interactive systems, and algorithm-driven platforms.
- Independent Optical Frequency Combs Powered 546 km Field Test of Twin-Field Quantum Key Distribution–
- –
- –
- –
- –
- Link Prediction Based on Smooth Evolution of Network Embedding–
A Smooth Evolution model for Network Embedding (called SENE) is proposed, which considers both deep-level features and temporal features to obtain the embedded representations of the network structure, and uses the transformer mechanism to effectively obtain the smooth evolution of network embedding.
- –
- –
- –
- –
- Using Game Theory to Classify Wireless Ad Hoc Network Attacks with Analysis on Countermeasures–
A classification which associates attack characteristics with types of game players and then examines the relationship between attack scenarios and types of corresponding game models is presented.
- Corpus-based English Vocabulary Teaching–
The present study studies the application of corpus into vocabulary teaching, including its advantages and ways, aimed at showing its value and function in the vocabulary teaching.
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.