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
- Jun SatoSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works86 from public data
- GPT4AIGChip: Towards Next-Generation AI Accelerator Design Automation via Large Language Models142
This work develops a framework called GPT4AIGChip, which features an automated demo-augmented prompt-generation pipeline utilizing in-context learning to guide LLMs towards creating high-quality AI accelerator design, and is the first to demonstrate an effective pipeline for LLM-powered automated AI accelerator generation.
- Instant-3D: Instant Neural Radiance Field Training Towards On-Device AR/VR 3D Reconstruction120
Instant-3D is proposed, an algorithm-hardware co-design acceleration framework that achieves instant on-device NeRF training and has enabled instant 3D reconstruction for AR/VR, requiring a reconstruction time of only 1.6 seconds per scene and meeting the AR/ VR power consumption constraint.
- BNS-GCN: Efficient Full-Graph Training of Graph Convolutional Networks with Partition-Parallelism and Random Boundary Node Sampling109
This work first analyzes why distributed GCN training is ineffective and identifies the underlying cause to be the excessive number of boundary nodes of each partitioned subgraph, which easily explodes the memory and communication costs for GCNTraining.
- Swift Parameter-free Attention Network for Efficient Super-Resolution106
The Swift Parameter-free Attention Network (SPAN), a highly efficient SISR model that balances parameter count, inference speed, and image quality, is proposed, showing that it outperforms existing efficient super-resolution models in terms of both image quality and inference speed.
- PipeGCN: Efficient Full-Graph Training of Graph Convolutional Networks with Pipelined Feature Communication102
This work not only provides a theoretical convergence analysis but also finds the convergence rate of PipeGCN to be close to that of the vanilla distributed GCN training without any staleness, and develops a smoothing method to further improve Pipe GCN's convergence.
- Towards Cognitive AI Systems: a Survey and Prospective on Neuro-Symbolic AI59
A systematic review of recent progress in NSAI is provided and the performance characteristics and computational operators of NSAI models are analyzed to discuss the challenges and potential future directions of NSAI.
- Web Service Selection with Incomplete or Inconsistent User Preferences47
A system to facilitate Web service selection in face of incomplete or inconsistent user preferences is presented, which utilizes the information of historical users to amend the active user's preference, so as to improve the results of service selection.
- A Survey on Graph Neural Network Acceleration: Algorithms, Systems, and Customized Hardware43
This survey provides a taxonomy of GNN acceleration, review the existing approaches, and suggest future research directions, and sets the stage for further development in this area.
- 39
- 35
- Gaussian Blending Unit: An Edge GPU Plug-in for Real-Time Gaussian-Based Rendering in AR/VR34
A Gaussian Blending Unit (GBU), an edge GPU plug-in module for real-time rendering in AR/VR applications that enhances rendering speed with a dedicated rendering engine that balances the workload across rows by aggregating computations from multiple Gaussians.
- Towards Efficient Neuro-Symbolic AI: From Workload Characterization to Hardware Architecture32
This paper systematically categorizes neuro-symbolic AI algorithms, and experimentally evaluates them in terms of runtime, memory, computational operators, sparsity, and system characteristics on CPUs, GPUs, and edge SoCs, revealing that neuro-symbolic models suffer from inefficiencies on off-the-shelf hardware.
- Machine-Learning and Radiomics-Based Preoperative Prediction of Ki-67 Expression in Glioma Using MRI Data31
This study introduced a novel approach for classifying Ki67 expression levels using MRI, which has been proven to be highly effective and demonstrated its potential in signalling a promising avenue for future research.
- Deep-learning and conventional radiomics to predict IDH genotyping status based on magnetic resonance imaging data in adult diffuse glioma29
An optimal machine-learning model for prediction of IDH genotyping is identified by combining deep-learning (DL) signatures and conventional radiomics (CR) features as model predictors by combining the LASSO–selected CR features and integrated DL signatures as model Predictors.
- DepthShrinker: A New Compression Paradigm Towards Boosting Real-Hardware Efficiency of Compact Neural Networks29
This work opens up a new compression paradigm for developing real-hardware efficient DNNs, leading to boosted hardware efficiency while maintaining model accuracy, and proposes a framework dubbed DepthShrinker, which develops hardware-friendly compact networks via shrinking the basic building blocks of existing efficient Dnns that feature irregular computation patterns into dense ones with much improved hardware utilization and thus real- hardware efficiency.
- Preference recommendation for personalized search28
This article proposes a preference recommendation scheme for complementing a user's CP-nets, so as to improve the accuracy of personalized search and conducts extensive simulation and user study to demonstrate the effectiveness.
- 3D-Carbon: An Analytical Carbon Modeling Tool for 3D and 2.5D Integrated Circuits27
3D-Carbon is proposed, an analytical carbon modeling tool designed to quantify the carbon emissions of 3D/2.5D ICs throughout their life cycle, and lays the initial foundation for future innovations in developing environmentally sustainable 3D/2.5D ICs.
- Deep RAW Image Super-Resolution. A NTIRE 2024 Challenge Survey21
The NTIRE 2024 RAW Image Super-Resolution Challenge is reviewed, highlighting the proposed solutions and results, and providing a gauge for the current state-of-the-art in RAW Image Super-Resolution.
- 18
- Unveiling the Evolution of Virtual Reality in Medicine: A Bibliometric Analysis of Research Hotspots and Trends over the Past 12 Years17
The number of publications on VR applications in the medical field has been steadily increasing year by year, and the USA is the leading country in this area, while the University of London stands out as the most published, and most influential institution.
- 17
- Deep learning based prediction of depression and anxiety in patients with type 2 diabetes mellitus using regional electronic health records15
The REDAPM effectively leverages the heterogeneous characteristics of regional EHR data, demonstrating strong predictive performance for depression onset in diabetic patients and shows potential for discovering significant clinical features, indicating considerable promise for clinical utility.
- 15
- 14
- EyeCoD14
This work proposes, develops, and validate a lensless FlatCambased eye tracking algorithm and accelerator co-design framework dubbed EyeCoD to enable eye tracking systems with a much reduced form-factor and boosted system efficiency without sacrificing the tracking accuracy, paving the way for next-generation eye tracking solutions.
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 2026-10-14. 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.