logo EDITE Natacha RUCHAUD
Identité
Natacha RUCHAUD
État académique
Thèse en cours...
Sujet: Video surveillance et vie privée
Direction de thèse:
Laboratoire:
Voisinage
Ellipse bleue: doctorant, ellipse jaune: docteur, rectangle vert: permanent, rectangle jaune: HDR. Trait vert: encadrant de thèse, trait bleu: directeur de thèse, pointillé: jury d'évaluation à mi-parcours ou jury de thèse.
Productions scientifiques
oai:hal.archives-ouvertes.fr:hal-01367568
PRIVACY PROTECTING, INTELLIGIBILITY PRESERVING VIDEO SURVEILLANCE
International audience
Video surveillance is increasingly omnipresent in our everyday life and is a key component of many security systems. Not only is the increasing number of cameras, but also the resolution of visual sensors and the performance of video processing algorithms. This evolution generates some important privacy concerns. This article introduces a new visual filter that includes a good trade-off between privacy and intelligi-bility. It ensures that people are unrecognizable while keeping the scene understandable in terms of events which allows machines to detect abnormal behavior. The algorithm operates in the DCT domain to be compliant with the popular JPEG and MPEG codecs. For each sensitive area of the picture (i.e. area where privacy needs to be protected), the proposed algorithm uses the low-frequency coefficients of the DCT to display a privacy preserved image of the region and the high-frequency coefficients to hide most of the original information. Finally, our process allows authorized users to nearly reverse the process thanks to the hidden information.
ICME 2016 IEEE International Conference on Multimedia and Expo https://hal.archives-ouvertes.fr/hal-01367568 ICME 2016 IEEE International Conference on Multimedia and Expo , Jul 2016, Seattle, United StatesARRAY(0x7f54708d3810) 2016-07-12
oai:hal.archives-ouvertes.fr:hal-01367565
Automatic Face Anonymization in Visual Data: Are we really well protected?
International audience
With the proliferation of digital visual data in diverse domains (video surveillance, social networks, medias, etc.), privacy concerns increase. Obscuring faces in images and videos is one option to preserve privacy while keeping a certain level of quality and intelligibility of the video. Most popular filters are blackener (black masking), pixelization and blurring. Even if it appears efficient at first sight, in terms of human perception, we demonstrate in this article that as soon as the category and the strength of the filter used to obscure faces can be (automatically) identified, there exist in the literature ad-hoc powerful approaches enable to partially cancel the impact of such filters with regards to automatic face recognition. Hence, evaluation is expressed in terms of face recognition rate associated with clean, obscured and de-obscured face images. Figure 1: Respectively, " 20 minutes " a French magazine using pixelization filter, " crimes " a French program using blurring filter and Street view by google using blurring filter.
Electronic Imaging https://hal.archives-ouvertes.fr/hal-01367565 Electronic Imaging, Feb 2016, San francisco, United StatesARRAY(0x7f54705859f0) 2016-02-15
oai:hal.archives-ouvertes.fr:hal-01367560
Privacy Protection Filter Using StegoScrambling in Video Surveillance
International audience
This paper introduces a new privacy filter adopted in the context of the DPT (Drone Protect Task) at MediaEval Benchmark 2015. Our proposed filter protects privacy by visually replacing sensitive RofI (Regions of Interest) by its shapes. A combination of steganography and scrambling is used in order to make this filter. Once the scrambling is applied on the pixels of the RofI, its MSB (Most Significant Bit) are hidden in the LSB (Least Significant Bit) of a cover image. Our filter fulfils four criteria defined by DPT: near-lossless reversibility, intelligibility, appropriateness and anonymization. We benchmarked the filter on the last three criteria and we get good results: 40 % for intelligibility and appropriateness, and 60 % for anonymization.
MediaEval https://hal.archives-ouvertes.fr/hal-01367560 MediaEval, Sep 2015, Wurzen, GermanyARRAY(0x7f546ffee920) 2015-09-09
oai:hal.archives-ouvertes.fr:hal-01367561
The impact of privacy protection filters on gender recognition
International audience
Deep learning-based algorithms have become increasingly efficient in recognition and detection tasks, especially when they are trained on large-scale datasets. Such recent success has led to a speculation that deep learning methods are comparable to or even outperform human visual system in its ability to detect and recognize objects and their features. In this paper, we focus on the specific task of gender recognition in images when they have been processed by privacy protection filters (e.g., blurring, masking, and pixelization) applied at different strengths. Assuming a privacy protection scenario, we compare the performance of state of the art deep learning algorithms with a subjective evaluation obtained via crowdsourcing to understand how privacy protection filters affect both machine and human vision.
SPIE Optical Engineering+ Applications https://hal.archives-ouvertes.fr/hal-01367561 SPIE Optical Engineering+ Applications, Aug 2015, San diego, United StatesARRAY(0x7f5470593398) 2015-08-28
oai:hal.archives-ouvertes.fr:hal-01367540
Efficient Privacy Protection in Video Surveillance by StegoScrambling
International audience
This paper presents a near lossless reversible system which allows a user to protect privacy in video surveillance (or still images) by replacing sensitive RofIs (Regions of interest) by its edges using steganography while keeping visual quality needed for security almost in real time. Our proposed system outperforms the state-of-the-art methods by fulfilling four criteria which are near lossless Reversible, Fast Computation (integrated in real time), Usability (preserves shape and motion of people to still recognize events) and Privacy protection (no more possibility to recognize people). The effectiveness of the proposed filter has been demonstrated using some face recognition algorithms on different images.
WIFS 7th IEEE International Workshop on Information Forensics and Security https://hal.archives-ouvertes.fr/hal-01367540 WIFS 7th IEEE International Workshop on Information Forensics and Security, Nov 2015, Rome, ItalyARRAY(0x7f5470578538) 2015-11-16
oai:hal.archives-ouvertes.fr:hal-01588279
ASePPI, an Adaptive Scrambling enabling Privacy Protection and Intelligibility in H.264/AVC
International audience
—The usage of video surveillance systems increases more and more every year and protecting people privacy becomes a serious concern. In this paper, we present ASePPI, an Adaptive Scrambling enabling Privacy Protection and Intelligibility. It operates in the DCT domain within the H.264 standard. For each residual block of the luminance channel inside the region of interest, we encrypt the coefficients. Whereas encrypted coefficients appear as noise in the protected image, the DC value is dedicated to restore some of the original information. Thus, the proposed approach automatically adapts the level of protection according to the resolution of the region of interest. Comparing to existing methods, our framework provides better privacy protection with some flexibilities on the appearance of the protected version yielding better visibility of the scene for monitoring. Moreover, the impact on the source coding stream is negligible. Indeed, the results demonstrate a slight decrease in the quality of the reconstructed images and a small percentage of bits overhead.
GRETSI https://hal.archives-ouvertes.fr/hal-01588279 GRETSI, Sep 2017, Juans les pins, FranceARRAY(0x7f5470581b40) 2017-09-05
oai:hal.archives-ouvertes.fr:hal-01588273
ASePPI: Robust Privacy Protection Against De-Anonymization Attacks
International audience
The evolution of the video surveillance systems generates questions concerning protection of individual privacy. In this paper, we design ASePPI, an Adaptive Scrambling enabling Privacy Protection and Intelligibility method operating in the H.264/AVC stream with the aim to be robust against de-anonymization attacks targeting the restoration of the original image and the re-identification of people. The proposed approach automatically adapts the level of protection according to the resolution of the region of interest. Compared to existing methods, our framework provides a better trade-off between the privacy protection and the visibility of the scene with robustness against de-anonymization attacks. Moreover, the impact on the source coding stream is negligible.
CVPR, Computer Vision and Pattern Recognition CVPR, Computer Vision and Pattern Recognition https://hal.archives-ouvertes.fr/hal-01588273 CVPR, Computer Vision and Pattern Recognition, Jul 2017, Honolulu, United States. CVPR, Computer Vision and Pattern Recognition, pp.1352 - 1359, 2017, <10.1109/CVPRW.2017.177>ARRAY(0x7f54707198b0) 2017-07-21
oai:hal.archives-ouvertes.fr:hal-01588265
De-genderization by body contours reshaping
International audience
This paper deals with privacy protection in video surveillance. More specifically, the main goal of this work is to make the gender of people no more recognizable while preserving enough information concerning body shape and motion of people for action classification. We denote this processing as de-genderization. Regarding the current state-of-art methods, most of them have privacy filters only dedicated to de-identify people. These methods do not automatically imply the suppression of visual semantic traits such as gender. Therefore, we propose two approaches that modify the visual appearance of the body shape in order to de-genderize people while keeping the possibility to interpret the video. In both methods we start by extracting the contour points attached to the body shape of people in videos. Then we either mix the coordinates of the body shape and a predefined model, or we smooth the body shape by successive polygonal approximations based on convexity. Our results demonstrate that both proposed approaches protect the gender information while preserving the global body movement. The second approach based on convexity better preserves the visibility of human activities.
2017 IEEE International Conference on Identity, Security and Behavior Analysis (ISBA) https://hal.archives-ouvertes.fr/hal-01588265 2017 IEEE International Conference on Identity, Security and Behavior Analysis (ISBA), 2017, pp.1 - 6. <10.1109/ISBA.2017.7947709>ARRAY(0x7f546ac32110) 2017
oai:hal.archives-ouvertes.fr:hal-01588276
ASePPI, an Adaptive Scrambling enabling Privacy Protection and Intelligibility in H.264/AVC
International audience
—The usage of video surveillance systems increases more and more every year and protecting people privacy becomes a serious concern. In this paper, we present ASePPI, an Adaptive Scrambling enabling Privacy Protection and Intelligibility. It operates in the DCT domain within the H.264 standard. For each residual block of the luminance channel inside the region of interest, we encrypt the coefficients. Whereas encrypted coefficients appear as noise in the protected image, the DC value is dedicated to restore some of the original information. Thus, the proposed approach automatically adapts the level of protection according to the resolution of the region of interest. Comparing to existing methods, our framework provides better privacy protection with some flexibilities on the appearance of the protected version yielding better visibility of the scene for monitoring. Moreover, the impact on the source coding stream is negligible. Indeed, the results demonstrate a slight decrease in the quality of the reconstructed images and a small percentage of bits overhead.
Eusipco https://hal.archives-ouvertes.fr/hal-01588276 Eusipco, Aug 2017, Kos, GreeceARRAY(0x7f5470589c60) 2017-08-28