Kurt Keutzer

47

Papers

447

Total Citations

Papers (47)

SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference

Why Do Multi-Agent LLM Systems Fail?

NeurIPS 2025arXiv

DeSiRe-GS: 4D Street Gaussians for Static-Dynamic Decomposition and Surface Reconstruction for Urban Driving Scenes

Efficient Deweahter Mixture-of-Experts with Uncertainty-Aware Feature-Wise Linear Modulation

K-Sort Arena: Efficient and Reliable Benchmarking for Generative Models via K-wise Human Preferences

UniDrive: Towards Universal Driving Perception Across Camera Configurations

A Lesson in Splats: Teacher-Guided Diffusion for 3D Gaussian Splats Generation with 2D Supervision

FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search

Trust Region Based Adversarial Attack on Neural Networks

ZeroQ: A Novel Zero Shot Quantization Framework

Prototypical Cross-Domain Self-Supervised Learning for Few-Shot Unsupervised Domain Adaptation

Learning Invariant Representations and Risks for Semi-Supervised Domain Adaptation

SelfAugment: Automatic Augmentation Policies for Self-Supervised Learning

NoisyQuant: Noisy Bias-Enhanced Post-Training Activation Quantization for Vision Transformers

Open-Vocabulary Point-Cloud Object Detection Without 3D Annotation

HAWQ: Hessian AWare Quantization of Neural Networks With Mixed-Precision

Domain Randomization and Pyramid Consistency: Simulation-to-Real Generalization Without Accessing Target Domain Data

Region Similarity Representation Learning

Visual Transformers: Where Do Transformers Really Belong in Vision Models?

NeRF-Det: Learning Geometry-Aware Volumetric Representation for Multi-View 3D Object Detection

Q-Diffusion: Quantizing Diffusion Models

SparseFusion: Fusing Multi-Modal Sparse Representations for Multi-Sensor 3D Object Detection

QD-BEV : Quantization-aware View-guided Distillation for Multi-view 3D Object Detection

SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation

MTTrans: Cross-Domain Object Detection with Mean Teacher Transformer

Image2Point: 3D Point-Cloud Understanding with 2D Image Pretrained Models

PreTraM: Self-Supervised Pre-training via Connecting Trajectory and Map

Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learning

Segment Any Motion in Videos

Bridging Viewpoint Gaps: Geometric Reasoning Boosts Semantic Correspondence

StreamDiffusion: A Pipeline-level Solution for Real-Time Interactive Generation

PromptCoT: Align Prompt Distribution via Adapted Chain-of-Thought

FireCaffe: Near-Linear Acceleration of Deep Neural Network Training on Compute Clusters

Shift: A Zero FLOP, Zero Parameter Alternative to Spatial Convolutions

Hessian-based Analysis of Large Batch Training and Robustness to Adversaries

ANODEV2: A Coupled Neural ODE Framework

Multi-source Domain Adaptation for Semantic Segmentation

Boundary thickness and robustness in learning models

HAWQ-V2: Hessian Aware trace-Weighted Quantization of Neural Networks

NovelD: A Simple yet Effective Exploration Criterion

Squeezeformer: An Efficient Transformer for Automatic Speech Recognition

K-LITE: Learning Transferable Visual Models with External Knowledge

A Fast Post-Training Pruning Framework for Transformers

Speculative Decoding with Big Little Decoder

Large Language Models are Visual Reasoning Coordinators

Towards Foundation Models for Scientific Machine Learning: Characterizing Scaling and Transfer Behavior

Regret Minimization for Partially Observable Deep Reinforcement Learning