paper-with-me

Papers

Dataset Distillation with Convexified Implicit Gradients

2023-02-13 · Noel Loo, Ramin Hasani, Mathias Lechner, Daniela Rus

We propose a new dataset distillation algorithm using reparameterization and convexification of implicit gradients (RCIG), that substantially improves the state-of-the-art. To this end, we first formulate dataset distillation as a bi-level optimization problem. Then, we show how implicit gradients can be effectively used to compute meta-gradient updates. We further equip the algorithm with a convexified approximation that corresponds to learning on top of a frozen finite-width neural tangent kernel. Finally, we improve bias in implicit gradients by parameterizing the neural network to enable analytical computation of final-layer parameters given the body parameters. RCIG establishes the new state-of-the-art on a diverse series of dataset distillation tasks. Notably, with one image per class, on resized ImageNet, RCIG sees on average a 108\% improvement over the previous state-of-the-art distillation algorithm. Similarly, we observed a 66\% gain over SOTA on Tiny-ImageNet and 37\% on CIFAR-100.

📄 PDF Abstract BibTeX arXiv:2302.06755

Code (2)

yolky/rcig 공식 구현 jax
Guang000/Awesome-Dataset-Distillation

Tasks

Dataset DistillationDataset Distillation - 1IPC

Similar Papers 제목 키워드 기반

Towards Stable and Storage-efficient Dataset Distillation: Matching Convexified Trajectory

2024-06-28 · CVPR 2025 1 · Wenliang Zhong, Haoyu Tang, Qinghai Zheng, Mingzhu Xu 외

The rapid evolution of deep learning and large language models has led to an exponential growth in the demand for training data, prompting the development of Dataset Distillation methods to address the challenges of mana…

Dataset Distillation

DreamHOI: Subject-Driven Generation of 3D Human-Object Interactions with Diffusion Priors

2024-09-12 · Thomas Hanwen Zhu, Ruining Li, Tomas Jakab

We present DreamHOI, a novel method for zero-shot synthesis of human-object interactions (HOIs), enabling a 3D human model to realistically interact with any given object based on a textual description. This task is comp…

Human-Object Interaction DetectionNeRF

Guiding Distribution Matching Distillation with Gradient-Based Reinforcement Learning

2026-04-21 · Linwei Dong, Ruoyu Guo, Ge Bai, Zehuan Yuan 외 arxiv

Diffusion distillation, exemplified by Distribution Matching Distillation (DMD), has shown great promise in few-step generation but often sacrifices quality for sampling speed. While integrating Reinforcement Learning (R…

Reinforcement Learning

Convexified Convolutional Neural Networks

2016-09-04 · ICML 2017 8 · Yuchen Zhang, Percy Liang, Martin J. Wainwright

We describe the class of convexified convolutional neural networks (CCNNs), which capture the parameter sharing of convolutional neural networks in a convex manner. By representing the nonlinear convolutional filters as …

Denoising

Understanding the Role of the Projector in Knowledge Distillation

2023-03-20 · Roy Miles, Krystian Mikolajczyk

In this paper we revisit the efficacy of knowledge distillation as a function matching and metric learning problem. In doing so we verify three important design decisions, namely the normalisation, soft maximum function,…

image-classificationImage ClassificationKnowledge DistillationMetric Learning+2