ELASTIC: Improving CNNs with Dynamic Scaling Policies

2018-12-13 · CVPR 2019 6 · Huiyu Wang, Aniruddha Kembhavi, Ali Farhadi, Alan Yuille, Mohammad Rastegari

Scale variation has been a challenge from traditional to modern approaches in computer vision. Most solutions to scale issues have a similar theme: a set of intuitive and manually designed policies that are generic and fixed (e.g. SIFT or feature pyramid). We argue that the scaling policy should be learned from data. In this paper, we introduce ELASTIC, a simple, efficient and yet very effective approach to learn a dynamic scale policy from data. We formulate the scaling policy as a non-linear function inside the network's structure that (a) is learned from data, (b) is instance specific, (c) does not add extra computation, and (d) can be applied on any network architecture. We applied ELASTIC to several state-of-the-art network architectures and showed consistent improvement without extra (sometimes even lower) computation on ImageNet classification, MSCOCO multi-label classification, and PASCAL VOC semantic segmentation. Our results show major improvement for images with scale challenges. Our code is available here: https://github.com/allenai/elastic

📄 PDF Abstract BibTeX arXiv:1812.05262

Code (1)

allenai/elastic 공식 구현 pytorch

Tasks

General Classification Multi-Label Classification MUlTI-LABEL-ClASSIFICATION Semantic Segmentation

Methods 이 논문이 사용한 방법론

Residual Connection 설명 없음

Convolution A convolution is a type of matrix operation, consisting of a kernel, a small matrix of weights, that slides over input data performing element-wise multiplication with the…

Average Pooling 설명 없음

Concatenated Skip Connection A Concatenated Skip Connection is a type of skip connection that seeks to reuse features by concatenating them to new layers, allowing more information to be retained from…

Global Average Pooling Global Average Pooling is a pooling operation designed to replace fully connected layers in classical CNNs. The idea is to generate one feature map for each corresponding…

Grouped Convolution A Grouped Convolution uses a group of convolutions - multiple kernels per layer - resulting in multiple channel outputs per layer. This leads to wider networks helping a…

1x1 Convolution A 1 x 1 Convolution is a convolution with some special properties in that it can be used for dimensionality reduction,…

Batch Normalization 설명 없음

ELASTIC: Efficiently Learning to Adaptively Scale Test-Time Compute for Generative Control Policies

2026-06-30 · Andrew Zou Li, Gokul Swamy, Yonatan Bisk, Andrea Bajcsy arxiv

Generative control policies (GCPs), such as diffusion policies and flow-based vision-language-action models, enable test-time scaling in robot control. Test-time compute can be allocated along two axes: sequential scalin…

Reinforcement Learning Robot Manipulation

Elastic Gossip: Distributing Neural Network Training Using Gossip-like Protocols

2018-12-06 · Siddharth Pramod

Distributing Neural Network training is of particular interest for several reasons including scaling using computing clusters, training at data sources such as IOT devices and edge servers, utilizing underutilized resour…

HFX: Joint Design of Algorithms and Systems for Multi-SLO Serving and Fast Scaling

2025-08-21 · Zahra Yousefijamarani, Xinglu Wang, Qian Wang, Morgan Lindsay Heisler 외 arxiv

Large language model (LLM) serving faces the dual challenge of meeting strict user-specific service-level objectives (SLOs) while minimizing computational cost under dynamic, multi-task workloads. Existing approaches eit…

A Hamiltonian Higher-Order Elasticity Framework for Dynamic Diagnostics(2HOED)

2025-04-29 · Ngueuleweu Tiwang Gildas

Machine learning detects patterns, block chain guarantees trust and immutability, and modern causal inference identifies directional linkages, yet none alone exposes the full energetic anatomy of complex systems; the Ham…

Anatomy Causal Inference Epidemiology

Task Generalization with Stability Guarantees via Elastic Dynamical System Motion Policies

2023-09-05 · Tianyu Li, Nadia Figueroa

Dynamical System (DS) based Learning from Demonstration (LfD) allows learning of reactive motion policies with stability and convergence guarantees from a few trajectories. Yet, current DS learning techniques lack the fl…