Enrique Ortiz

Face Recognition in Movie Trailers via Mean Sequence Sparse Representation-Based Classification

This paper presents an end-to-end video face recognition system, addressing the difficult problem of identifying a video face track using a large dictionary of still face images of a few hundred people, while rejecting unknown individuals. A straightforward application of the popular l1-minimization for face recognition on a frame-by-frame basis is prohibitively expensive, so we propose a novel algorithm Mean Sequence SRC (MSSRC) that performs video face recognition using a joint optimization leveraging all of the available video data and the knowledge that the face track frames belong to the same individual. By adding a strict temporal constraint to the l1-minimization that forces individual frames in a face track to all reconstruct a single identity, we show the optimization reduces to a single minimization over the mean of the face track. We also introduce a new Movie Trailer Face Dataset collected from 101 movie trailers on YouTube. Finally, we show that our method matches or outperforms the state-of-the-art on three existing datasets (YouTube Celebrities, YouTube Faces, and Buffy) and our unconstrained Movie Trailer Face Dataset. More importantly, our method excels at rejecting unknown identities by at least 8% in average precision.

Face recognition for web-scale datasets

With millions of users and billions of photos, web-scale face recognition is a challenging task that demands speed, accuracy, and scalability. Most current approaches do not address and do not scale well to Internet-sized scenarios such as tagging friends or finding celebrities. Focusing on web-scale face identification, we gather an 800,000 face dataset from the Facebook social network that models real-world situations where specific faces must be recognized and unknown identities rejected. We propose a novel Linearly Approximated Sparse Representation-based Classification (LASRC) algorithm that uses linear regression to perform sample selection for @?^1-minimization, thus harnessing the speed of least-squares and the robustness of sparse solutions such as SRC. Our efficient LASRC algorithm achieves comparable performance to SRC with a 100-250 times speedup and exhibits similar recall to SVMs with much faster training. Extensive tests demonstrate our proposed approach is competitive on pair-matching verification tasks and outperforms current state-of-the-art algorithms on open-universe identification in uncontrolled, web-scale scenarios.

Evaluation of face recognition techniques for application to facebook

This paper evaluates face recognition applied to the real-world application of Facebook. Because papers usually present results in terms of accuracy on constrained face datasets, it is difficult to assess how they would work on natural data in a real-world application. We present a method to automatically gather and extract face images from Facebook, resulting in over 60,000 faces datasets, we evaluate a variety of well-known face recognition algorithms (PCA, LDA, ICA, SVMs) against holistic performance metrics of accuracy, speed, memory usage, and storage size. SVMs perform best with ~65% accuracy, but lower accuracy algorithms such as IPCA are orders of magnitude more efficient in memory consumption and speed, yielding a more feasible system.

A Scalable and Efficient Outlier Detection Strategy for Categorical Data

Outlier detection has received significant attention in many applications, such as detecting credit card fraud or network intrusions. Most existing research focuses on numerical datasets, and cannot directly apply to categorical sets where there is little sense in calculating distances among data points. Furthermore, a number of outlier detection methods require quadratic time with respect to the dataset size and usually multiple dataset scans. These characteristics are undesirable for large datasets, potentially scattered over multiple distributed sites. In this paper, we introduce Attribute Value Frequency (A VF), a fast and scalable outlier detection strategy for categorical data. A VF scales linearly with the number of data points and attributes, and relies on a single data scan. AVF is compared with a list of representative outlier detection approaches that have not been contrasted against each other. Our proposed solution is experimentally shown to be significantly faster, and as effective in discovering outliers.

Evaluating Open-Universe Face Identification on the Web

Face recognition is becoming a widely used technique to organize and tag photos. Whether searching, viewing, or organizing photos on the web or in personal photo albums, there is a growing demand to index real-world photos by the subjects in them. Even consumer platforms such as Google Picasa, Microsoft Photo Gallery, and social network sites such as Facebook have integrated forms of automated face tagging and recognition, furthermore, a number of libraries and cloud-based APIs that perform face recognition have become available. With such a plethora of choices, comparisons of recent advances become more important to gauge the state of progress in the field. This paper evaluates face identification in the context of not only research algorithms, but also considers consumer photo products, client-side libraries, and cloud-based APIs on a new, large-scale dataset derived from PubFig83 and LFW in a realistic open-universe scenario.

Effects of Logo Programming on Understanding Variables

This study investigated whether there were significant differences in understanding the concept of variable and in attitudes toward mathematics among sixth-grade students (n = 89) who used a Logo graphics approach, students who used a textbook-based approach, and students who received no instruction on the concept of variable. A posttest of variable understanding (UCVI) was administered immediately and three weeks after instruction was completed. In addition, the Robustness Semantic Differential (an attitudinal survey) was given at the end of the instructional treatment. For the UCVI immediate test, the Logo group scored higher (p < .01) than the control group, but there was no significant difference between the Logo and the textbook groups. However, for the UCVI delayed retention test, the Logo group scored significantly higher (p < .01) than both the textbook and control groups. Analysis of the attitudinal survey indicated students had more positive attitudes toward computer-related aspects of instruction.