Sign Function

The sign function, a simple yet powerful mathematical operation, plays a crucial role in various scientific domains, particularly in signal processing, machine learning, and computer vision. Current research focuses on refining its application within algorithms like projected gradient descent (PGD) for adversarial attacks on neural networks, and exploring alternatives to improve efficiency and effectiveness, such as using raw gradients or dynamic sign functions tailored to specific data. These advancements aim to enhance the accuracy and speed of computations, impacting fields ranging from image retrieval to differentiable matrix operations in computer vision.

Papers

March 30, 2024

Extracting Manifold Information from Point Clouds
Patrick Guidotti
Point Cloud Class Manifold Low Dimensional Manifold Small Kernel Flamelet Generated Manifold Sign Function

December 3, 2023

Rethinking PGD Attack: Is Sign Function Necessary?
Junjie Yang, Tianlong Chen, Xuxi Chen, Zhangyang Wang, Yingbin Liang
Adversarial Attack Gradient Descent Adversarial Input Multi Step Attack Projected Gradient Descent Attack Sign Function

October 14, 2022

Cross-Scale Context Extracted Hashing for Fine-Grained Image Binary Encoding
Xuetong Xue, Jiaying Shi, Xinxue He, Shenghui Xu, Zhaoming Pan
Fine Grained Image Retrieval Hash Function Deep Hashing Hash Learning Sign Function

January 29, 2022

Fast Differentiable Matrix Square Root and Inverse Square Root
Yue Song, Nicu Sebe, Wei Wang
Singular Value Decomposition Newton Iteration Polynomial Expansion Square Root Sign Function Differentiable Matrix

November 27, 2021

Adaptive Perturbation for Adversarial Attack
Zheng Yuan, Jie Zhang, Zhaoyan Jiang, Liangliang Li, Shiguang Shan
Adversarial Attack Adversarial Example Adversarial Perturbation Gradient Based Attack Exact Gradient Targeted Perturbation Sign Function

Sign Function

Papers

Extracting Manifold Information from Point Clouds

Rethinking PGD Attack: Is Sign Function Necessary?

Cross-Scale Context Extracted Hashing for Fine-Grained Image Binary Encoding

Fast Differentiable Matrix Square Root and Inverse Square Root

Adaptive Perturbation for Adversarial Attack