Towards privacy-preserving recognition of human activities

Dai, Ji; Saghafi, Behrouz; Wu, Q. M. Jonathan; Konrad, Janusz; Ishwar, Prakash

doi:10.1109/icip.2015.7351605

Cited by 39 publications

(32 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In (Ryoo et al 2017), it is demonstrated that reliable action recognition may be achieved at low resolutions by learning appropriate downsampling transformations. Furthermore, trade-offs between resolution and action recognition accuracy are discussed in (Dai et al 2015). Furthermore, (Oh, Fritz, and Schiele 2017) propose an adversarial method to learn the image perturbation so as to fool the face identity classifier, but the adversarial perturbed images are visually exposing the identity privacy.…”

Section: Visual Privacy Protectionmentioning

confidence: 99%

Adversarial Learning of Privacy-Preserving and Task-Oriented Representations

Xiao

Tsai

Sohn

et al. 2020

AAAI

View full text Add to dashboard Cite

Data privacy has emerged as an important issue as data-driven deep learning has been an essential component of modern machine learning systems. For instance, there could be a potential privacy risk of machine learning systems via the model inversion attack, whose goal is to reconstruct the input data from the latent representation of deep networks. Our work aims at learning a privacy-preserving and task-oriented representation to defend against such model inversion attacks. Specifically, we propose an adversarial reconstruction learning framework that prevents the latent representations decoded into original input data. By simulating the expected behavior of adversary, our framework is realized by minimizing the negative pixel reconstruction loss or the negative feature reconstruction (i.e., perceptual distance) loss. We validate the proposed method on face attribute prediction, showing that our method allows protecting visual privacy with a small decrease in utility performance. In addition, we show the utility-privacy trade-off with different choices of hyperparameter for negative perceptual distance loss at training, allowing service providers to determine the right level of privacy-protection with a certain utility performance. Moreover, we provide an extensive study with different selections of features, tasks, and the data to further analyze their influence on privacy protection.

show abstract

Section: Visual Privacy Protectionmentioning

confidence: 99%

Adversarial Learning of Privacy-Preserving and Task-Oriented Representations

Xiao

Tsai

Sohn

et al. 2020

AAAI

View full text Add to dashboard Cite

show abstract

“…Szegedy et al [23] investigated on deep convolutional networks with action recognition using pre-recorded videos. Leenes et al [24] studied on the privacy issues associated with data protection Dai et al [25] proposed a novel method towards human action recognition with privacy preserved. Kumar et al [5] explored deep learning algorithms and resolution images besides spatial relationships to recognize human actions.…”

Section: Related Workmentioning

confidence: 99%

Privacy preserving human activity recognition framework using an optimized prediction algorithm

Kumar¹,

Harikiran²

2022

IJ-AI

View full text Add to dashboard Cite

Human activity recognition, in computer vision research, is the area of growing interest as it has plethora of real-world applications. Inferring actions from one or more persons captured through a live video has its immense utility in the contemporary era. Same time, protecting privacy of humans is to be given paramount importance. Many researchers contributed towards this end leading to privacy preserving action recognition systems. However, having an optimized model that can withstand any adversary models that strives to disclose privacy information. To address this problem, we proposed an algorithm known optimized prediction algorithm for privacy preserving activity recognition (OPA-PPAR) based on deep neural networks. It anonymizes video content to have adaptive privacy model that defeats attacks from adversaries. The privacy model enhances the privacy of humans while permitting highly accurate approach towards action recognition. The algorithm is implemented to realize privacy preserving human activity recognition framework (PPHARF). The visual recognition of human actions is made using an underlying adversarial learning process where the anonymization is optimized to have an adaptive privacy model. A dataset named human metabolome database (HMDB51) is used for empirical study. Our experiments with using Python data science platform reveal that the OPA-PPAR outperforms existing methods.

show abstract

“…On the other hand, anonymized videos are intentionally captured or processed to be in special low quality conditions that only allow for the recognition of some target events or activities (Butler et al, 2015;Dai et al, 2015;Ryoo et al, 2017;Ren et al, 2018) as shown in Figure 8. And, Winkler et al (2014) introduced cartoon-like effects as shown in Figure 9.…”

Section: Privacy-preserving Approaches To Human and Human Activity Recognitionmentioning

confidence: 99%

A review of privacy-preserving human and human activity recognition

Jung

2020

International Journal on Smart Sensing and Intelligent Systems

View full text Add to dashboard Cite

Many automation technologies using software are making humans convenient. One of these technologies is to collect data through cameras and sensors that are common in personal life and automatically recognize human and human activities. The goal of automation is to analyze the various types of big data that are difficult to perform mechanical data mining. Raw data collected from cameras and sensors are nothing but big data before analysis. In this case, how to protect data by secure storage is the most important issue. However, when the context-aware semantic information such as a specific person and his behavior is extracted from the analysis, the security sensitivity is increased. In other words, the secondary information generated by interpreting and extracting personal location and behavioral information contained in images and videos is linked to other personal information, causing privacy infringement issues. Privacy issues become important because there is a lot of software that everyone can access. Therefore, it is necessary to study privacy protection methods in the automatic recognition of human and human activities. This paper analyzes the cutting-edge research trends, techniques, and issues of privacy-preserving human and human activity recognition.

show abstract

Towards privacy-preserving recognition of human activities

Cited by 39 publications

References 8 publications

Adversarial Learning of Privacy-Preserving and Task-Oriented Representations

Adversarial Learning of Privacy-Preserving and Task-Oriented Representations

Privacy preserving human activity recognition framework using an optimized prediction algorithm

A review of privacy-preserving human and human activity recognition

Contact Info

Product

Resources

About