Biometrics are widely used for authentication in consumer devices and business settings as they provide sufficiently strong security instant verification and convenience for users. However, biometrics are hard to keep secret, stolen biometrics pose lifelong security risks to users as they cannot be reset and re-issued, and transactions authenticated by biometrics across different systems are linkable and traceable back to the individual identity. In addition, their cost-benefit analysis does not include personal implications to users, who are least prepared for the imminent negative outcomes, and are not often given equally convenient alternative authentication options.

The impact of citizen journalism raises important video integrity and credibility issues. In this article, we introduce Vamos, the first user transparent video “liveness” verification solution based on video motion, that accommodates the full range of camera movements, and supports videos of arbitrary length. Vamos uses the agreement between video motion and camera movement to corroborate the video authenticity. Vamos can be integrated into any mobile video capture application without requiring special user training. We develop novel attacks that target liveness verification solutions. The attacks leverage both fully automated algorithms and trained human experts. We introduce the concept of video motion categories to annotate the camera and user motion characteristics of arbitrary videos. We show that the performance of Vamos depends on the video motion category. Even though Vamos uses motion as a basis for verification, we observe a surprising and seemingly counter-intuitive resilience against attacks performed on relatively “stationary” video chunks, which turn out to contain hard-to-imitate involuntary movements. We show that overall the accuracy of Vamos on the task of verifying whole length videos exceeds 93 percent against the new attacks.

The increasing popular interest in personal telemetry, also called the Quantified Self or “lifelogging”, has induced a popularity surge for wearable personal fitness trackers. Fitness trackers automatically collect sensor data about the user throughout the day, and integrate it into social network accounts. Solution providers have to strike a balance between many constraints, leading to a design process that often puts security in the back seat. Case in point, we reverse engineered and identified security vulnerabilities in Fitbit Ultra and Gammon Forerunner 610, two popular and representative fitness tracker products. We introduce FitBite and GarMax, tools to launch efficient attacks against Fitbit and Garmin. We devise SensCrypt, a protocol for secure data storage and communication, for use by makers of affordable and lightweight personal trackers. SensCrypt thwarts not only the attacks we introduced, but also defends against powerful JTAG Read attacks. We have built Sens.io, an Arduino Uno based tracker platform, of similar capabilities but at a fraction of the cost of current solutions. On Sens.io, SensCrypt imposes a negligible write overhead and significantly reduces the end-to-end sync overhead of Fitbit and Garmin.

The ubiquitous and connected nature of camera-equipped mobile devices has greatly increased the value and importance of visual information they capture. Today, broadcasting videos from camera phones uploaded by unknown users is admissible on news networks, and banking customers expect to be able to deposit checks using mobile devices. In this paper, we introduce Movee, a system that addresses the fundamental question of whether the visual stream uploaded by a user has been captured live on a mobile device, and has not been tampered with by an adversary. Movee leverages the mobile device motion sensors and the intrinsic user movements during the shooting of the video. Movee exploits the observation that the movement of the scene recorded on the video stream should be related to the movement of the device simultaneously captured by the accelerometer. Contrary to existing algorithms, Movee has the unique strength of not depending on the audio track. We introduce novel attacks that focus on Movee's defenses, to fabricate acceleration data that mimics the motion observed in targeted videos. We use smartphones and wearable smart glasses to collect both genuine and attack data from 13 users. Our experiments show that Movee is able to efficiently detect human and automatically generated plagiarized videos: Movee's accuracy ranges between 68-93 percent on a smartphone, and between 76-91 percent on a Google Glass device.

Citizen journalism videos increasingly complement or even replace the professional news coverage through direct reporting by event witnesses. This raises questions of the integrity and credibility of such videos. We introduce Vamos, the first user transparent video "liveness" verification solution based on video motion, that can be integrated into any mobile video capture application without requiring special user training. Vamos' algorithm not only accommodates the full range of camera movements, but also supports videos of arbitrary length. We develop strong attacks both by utilizing fully automated attackers and by employing trained human experts for creating fraudulent videos to thwart mobile video verification systems.

The increasing interest in personal telemetry has induced a popularity surge for wearable personal fitness trackers. Such trackers automatically collect sensor data about the user throughout the day, and integrate it into social network accounts. Solution providers have to strike a balance between many constraints, leading to a design process that often puts security in the back seat. Case in point, we reverse engineered and identified security vulnerabilities in Fit bit Ultra and Gammon Forerunner 610, two popular and representative fitness tracker products. We introduce Fit Bite and GarMax, tools to launch efficient attacks against Fit bit and Garmin. We devise SensCrypt, a protocol for secure data storage and communication, for use by makers of affordable and lightweight personal trackers. SensCrypt thwarts not only the attacks we introduced, but also defends against powerful JTAG Read attacks. We have built Sens.io, an Arduino Uno based tracker platform, of similar capabilities but at a fraction of the cost of current solutions. On Sens.io, SensCrypt imposes a negligible write overhead and significantly reduces the end-to-end sync overhead of Fit bit and Garmin.

The visual information captured with camera-equipped mobile devices has greatly appreciated in value and importance as a result of their ubiquitous and connected nature. Today, banking customers expect to be able to deposit checks using mobile devices, and broadcasting videos from camera phones uploaded by unknown users is admissible on news networks. We present Movee, a system that addresses the fundamental question of whether the visual stream coming into a mobile app from the camera of the device can be trusted to be un-tampered with, live data, before it can be used for a variety of purposes.

Computational cost of classification is as important as accuracy in on-line classification systems. The computational cost is usually dominated by the cost of computing implicit features of the raw input data. Very few efforts have been made to design classifiers which perform effectively with limited computational power; instead, feature selection is usually employed as a pre-processing step to reduce the cost of running traditional classifiers. We present CoCoST, a novel and effective approach for building classifiers which achieve state-of-the-art classification accuracy, while keeping the expected computational cost of classification low, even without feature selection. CoCost employs a wide range of novel cost-aware decision trees, each of which is tuned to specialize in classifying instances from a subset of the input space, and judiciously consults them depending on the input instance in accordance with a cost-aware meta-classifier. Experimental results on a network flow detection application show that, our approach can achieve better accuracy than classifiers such as SVM and random forests, while achieving 75%-90% reduction in the computational costs.

MOTIVATION: Co-evolution is a powerful mechanism for understanding protein function. Prior work in this area has shown that co-evolving proteins are more likely to share the same function than those that do not because of functional constraints. Many of the efforts founded on this observation, however, are at the level of entire sequences, implicitly assuming that the complete protein sequence follows a single evolutionary trajectory. Since it is well known that a domain can exist in various contexts, this assumption is not valid for numerous multi-domain proteins. Motivated by these observations, we introduce a novel technique called Coevolutionary-Matrix that captures co-evolution between regions of two proteins. Instead of using existing domain information, the method exploits residue-level conservation to identify co-evolving regions that might correspond to domains. RESULTS: We show that the Coevolutionary-Matrix method can detect greater number of known functional associations for the Escherichia coli proteins when compared with earlier implementations of phylogenetic profiles. Furthermore, co-evolving regions of proteins detected by our method enable us to make hypotheses about their specific functions, many of which are supported by existing biochemical studies.

Text-password based authentication schemes are a popular means of authenticating users in computer systems. Standard security practices that were intended to make passwords more difficult to crack, such as requiring users to have passwords that "look random" (high entropy), have made password systems less usable and paradoxically, less secure. In this work, we address the need for enhancing the usability of existing text-password systems without necessitating any modifications to the existing password authentication infrastructure. We propose, develop and evaluate a system that automatically generates memorable mnemonics for a given password based on a text-corpus. Initial experimental results suggest that automatic mnemonic generation is a promising technique for making text-password systems more usable. Our system was able to generate mnemonics for 80.5% of six-character passwords and 62.7% of seven-character passwords containing lower-case characters (a-z), even when the text-corpus size is extremely small (1000 sentences).

The increasing interest in multilingual applications like speech-to-speech translation systems is accompanied by the need for speech recognition front-ends in many languages that can also handle multiple input languages at the same time. We describe a universal speech recognition system that fulfills such needs. It is trained by sharing speech and text data across languages and thus reduces the number of parameters and overhead significantly at the cost of only slight accuracy loss. The final recognizer eases the burden of maintaining several monolingual engines, makes dedicated language identification obsolete and allows for code-switching within an utterance. To achieve these goals we developed new methods for constructing multilingual acoustic models and multilingual n-gram language models.