24
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Subliminal Probing for Private Information via EEG-Based BCI Devices

      Preprint

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Martinovic et al. proposed a Brain-Computer-Interface (BCI) -based attack in which an adversary is able to infer private information about a user, such as their bank or area-of-living, by analyzing the user's brain activities. However, a key limitation of the above attack is that it is intrusive, requiring user cooperation, and is thus easily detectable and can be reported to other users. In this paper, we identify and analyze a more serious threat for users of BCI devices. We propose a it subliminal attack in which the victim is attacked at the levels below his cognitive perception. Our attack involves exposing the victim to visual stimuli for a duration of 13.3 milliseconds -- a duration usually not sufficient for conscious perception. The attacker analyzes subliminal brain activity in response to these short visual stimuli to infer private information about the user. If carried out carefully, for example by hiding the visual stimuli within screen content that the user expects to see, the attack may remain undetected. As a consequence, the attacker can scale it to many victims and expose them to the attack for a long time. We experimentally demonstrate the feasibility of our subliminal attack via a proof-of-concept study carried out with 27 subjects. We conducted experiments on users wearing Electroencephalography-based BCI devices, and used portrait pictures of people as visual stimuli which were embedded within the background of an innocuous video for a time duration not exceeding 13.3 milliseconds. Our experimental results show that it is feasible for an attacker to learn relevant private information about the user, such as whether the user knows the identity of the person for which the attacker is probing.

          Related collections

          Most cited references10

          • Record: found
          • Abstract: found
          • Article: not found

          Contribution of human hippocampal region to novelty detection.

          The ability to respond to unexpected stimuli (the 'orienting response') is a fundamental characteristic of mammalian behaviour, but the brain mechanisms by which novelty is detected remain poorly defined. Electrophysiological recordings of scalp and intracranial event-related potentials (ERPs) have shown that novel stimuli activate a distributed network involving prefrontal and posterior association cortex. In addition, ERP and single-neuron recordings, as well as neuroimaging and modelling studies, have suggested that temporal cortical regions, including the hippocampus, are also involved. To examine further the role of the medial temporal lobe in novelty processing, I measured physiological responses to novel auditory and tactile stimuli in patients with damage to the posterior hippocampal region. In normal control subjects, unexpected novel stimuli produce a characteristic ERP signal, accompanied by an autonomic skin response. Both responses are reduced in hippocampal lesion patients, whereas the response to expected control stimuli is unaffected. Thus the hippocampal region, in addition to its known role in memory formation, is an essential component of the distributed limbic-cortical network that detects and responds to novel stimuli.
            Bookmark
            • Record: found
            • Abstract: found
            • Article: not found

            Event-related brain potentials differentiate positive and negative mood adjectives during both supraliminal and subliminal visual processing.

            This experiment provides brain event-related potential (ERP) evidence for differential processing of visually presented pleasant and unpleasant affectively valent words (mood adjectives) for both supraliminal (40 ms) and subliminal (unmasked, 1 ms) stimulus durations. Unpleasant words elicited a more positive amplitude than pleasant words in both durations. ERP components (P1, N1, P2, P3, and a late positive potential; LP) were measured at six electrode sites (F3, F4, P3, P4, CzPz, Oz). ERPs to subliminal stimuli demonstrated differences between pleasant and unpleasant words in the left hemisphere across all measured components. Supraliminal processing showed similar differences in the left hemisphere for early components (P1 and N1), but bilateral differences for late components (P3 and LP). Activity in the P2 time window was associated with the divergence between supraliminal and subliminal affective responses. Implications for the study of affect and consciousness are discussed.
              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Dissociations of face and object recognition in developmental prosopagnosia.

              Neuropsychological studies with patients suffering from prosopagnosia have provided the main evidence for the hypothesis that the recognition of faces and objects rely on distinct mechanisms. Yet doubts remain, and it has been argued that no case demonstrating an unequivocal dissociation between face and object recognition exists due in part to the lack of appropriate response time measurements (Gauthier et al., 1999). We tested seven developmental prosopagnosics to measure their accuracy and reaction times with multiple tests of face recognition and compared this with a larger battery of object recognition tests. For our systematic comparison, we used an old/new recognition memory paradigm involving memory tests for cars, tools, guns, horses, natural scenes, and houses in addition to two separate tests for faces. Developmental prosopagnosic subjects performed very poorly with the face memory tests as expected. Four of the seven prosopagnosics showed a very strong dissociation between the face and object tests. Systematic comparison of reaction time measurements for all tests indicates that the dissociations cannot be accounted for by differences in reaction times. Contrary to an account based on speed accuracy tradeoffs, prosopagnosics were systematically faster in nonface tests than in face tests. Thus, our findings demonstrate that face and nonface recognition can dissociate over a wide range of testing conditions. This is further support for the hypothesis that face and nonface recognition relies on separate mechanisms and that developmental prosopagnosia constitutes a disorder separate from developmental agnosia.
                Bookmark

                Author and article information

                Journal
                2013-12-20
                2017-05-29
                Article
                1312.6052
                8d410ebf-1ffc-452d-9c38-dd9a36588357

                http://arxiv.org/licenses/nonexclusive-distrib/1.0/

                History
                Custom metadata
                under review for a journal publication
                cs.CR

                Security & Cryptology
                Security & Cryptology

                Comments

                Comment on this article