1

PrivLBS: Local Differential Privacy for Location-Based Services with Staircase Randomized Response

We propose the first LDP framework for a variety of location-based services (namely ``PrivLBS’’), which privately collects and analyzes user locations with high utility

Han Wang, Hanbin Hong, Li Xiong, Zhan Qin, Yuan Hong

UniCR: Universally Approximated Certified Robustness via Randomized Smoothing

We propose the first randomized smoothing-based universal certified robustness approximation framework against any ℓp perturbations with any continuous noise PDF.

Hanbin Hong, Binghui Wang, Yuan Hong

UniCR: Universally Approximated Certified Robustness via Randomized Smoothing

An Eye for an Eye: Defending against Gradient-based Attacks with Gradients

In this paper, we show that the gradients can also be exploited as a powerful weapon to defend against adversarial attacks.

Hanbin Hong, Yuan Hong, Yu Kong

Rit-18: A novel dataset for compositional group activity understanding

In this paper, we propose a new large-scale untrimmed compositional group activity dataset RIT-18 based on the volleyball games captured from YouTube.

Junwen Chen, Haiting Hao, Hanbin Hong, Yu Kong

Privacy Attributes-aware Message Passing Neural Network for Visual Privacy Attributes Classification

We propose a novel end-to-end Privacy Attributes-aware Message Passing Neural Network (PA-MPNN) to address Visual Privacy Attribute Classification.

Hanbin Hong, Wentao Bao, Yuan Hong, Yu Kong