Abstract

We propose a framework for automatically generating multiple clips suitable for video editing by simulating pan-tilt-zoom camera movements within the frame of a single static camera. Assuming important actors and objects can be localized using computer vision techniques, our method requires only minimal user input to define the subject matter of each sub-clip. The composition of each sub-clip is automatically computed in a novel L1-norm optimization framework. Our approach encodes several common cinematographic practices into a single convex cost function minimization problem, resulting in aesthetically pleasing sub-clips which can easily be edited together using off-the-shelf multi-clip video editing software. We demonstrate our approach on five video sequences of a live theatre performance by generating multiple synchronized subclips for each sequence.

Abstract

Formant frequencies represent resonances of vocal tract system during the production of speech signals. Bandwidths associated with the formant frequencies are important parameters in analysis and synthesis of speech signals. In this paper, a method is proposed to extract the bandwidths associated with formant frequencies, by analysing short segments (2–3 ms) of speech signal. The method is based on two important properties of group delay function (GDF): (a) The GDF exhibits prominent peaks at resonant frequencies and (b) the influence of one resonant frequency on other resonances is negligible in GDF. The accuracy of the method is demonstrated for synthetic signals generated using all-pole filters. The method is evaluated by extracting bandwidths of synthetic signals in closed phase and open phase regions within a pitch period. The accuracy of the proposed method is also compared with that of two other methods, one based on linear prediction analysis of speech signals, and another based on filterbank arrays for obtaining amplitude envelopes and instantaneous frequency signals. Results indicate that the method based on the properties of GDF is suitable for accurate extraction of formant bandwidths, even from short segments of speech signal within a pitch period.

Abstract

Word sketches are one-page automatic, corpus-based summaries of a word’s grammatical and collocational behaviour. These are widely used for studying a language and in lexicography. Sketch Engine is a leading corpus tool which takes as input a corpus and generates word sketches for the words of that language. It also generates a thesaurus and ‘sketch differences’, which specify similarities and differences between near-synonyms. In this paper, we present the functionalities of Sketch Engine for Hindi. We collected HindiWaC, a web crawled corpus for Hindi with 240 million words. We lemmatized, POS tagged the corpus and then loaded it into Sketch Engine.

Abstract

In this paper, we present a new multistage approach for SfM reconstruction of a single component. Our method begins with building a coarse 3D reconstruction using highscale features of given images. This step uses only a fraction of features and is fast. We enrich the model in stages by localizing remaining images to it and matching and triangulating remaining features. Unlike traditional incremental SfM, localization and triangulation steps in our approach are made efficient and embarrassingly parallel using geometry of the coarse model. The coarse model allows us to use 3D-2D correspondences based direct localization techniques to register remaining images. We further utilize the geometry of the coarse model to reduce the pair-wise image matching effort as well as to perform fast guided feature matching for majority of features. Our method produces similar quality models as compared to incremental SfM methods while being notably fast and parallel. Our algorithm can reconstruct a 1000 images dataset in 15 hours using a single core, in about 2 hours using 8 cores and in a few minutes by utilizing full parallelism of about 200 cores.

Abstract

Optimally controlled initiation of intramolecular H-transfer in malonaldehyde is accomplished by designing a sequence of ultrashort (∼80 fs) down-chirped pump-dump ultra violet (UV)-laser pulses through an optically bright electronic excited [S2 (ππ∗)] state as a mediator. The sequence of such laser pulses is theoretically synthesized within the framework of optimal control theory (OCT) and employing the well-known pump-dump scheme of Tannor and Rice [D.J. Tannor, S.A. Rice, J. Chem. Phys. 83, 5013 (1985)]. In the OCT, the control task is framed as the maximization of cost functional defined in terms of an objective function along with the constraints on the field intensity and system dynamics. The latter is monitored by solving the time-dependent Schr¨odinger equation. The initial guess, laser driven dynamics and the optimized pulse structure (i.e., the spectral content and temporal profile) followed by associated mechanism involved in fulfilling the control task are examined in detail and discussed. A comparative account of the dynamical outcomes within the Condon approximation for the transition dipole moment versus its more realistic value calculated ab initio is also presented.

Abstract

We propose an approach for segmenting the individual buildings in typical skyline images. Our approach is based on a Markov Random Field (MRF) formulation that exploits the fact that such images contain overlapping objects of similar shapes exhibiting a “tiered” structure. Our contributions are the following: (1) A dataset of 120 highresolution skyline images from twelve different cities with over 4,000 individually labeled buildings that allows us to quantitatively evaluate the performance of various segmentation methods, (2) An analysis of low-level features that are useful for segmentation of buildings, and (3) A shapeconstrained MRF formulation that enforces shape priors over the regions. For simple shapes such as rectangles, our formulation is significantly faster to optimize than a standard MRF approach, while also being more accurate. We experimentally evaluate various MRF formulations and demonstrate the effectiveness of our approach in segmenting skyline images

Abstract

The performance of binary classification tasks, such as action classification and object detection, is often measured in terms of the average precision (AP). Yet it is common practice in computer vision to employ the support vector machine (SVM) classifier, which optimizes a surrogate 0-1 loss. The popularity of SVM can be attributed to its empirical performance. Specifically, in fully supervised settings, SVM tends to provide similar accuracy to the AP-SVM classifier, which directly optimizes an AP-based loss. However, we hypothesize that in the significantly more challenging and practically useful setting of weakly supervised learning, it becomes crucial to optimize the right accuracy measure. In order to test this hypothesis, we propose a novel latent AP-SVM that minimizes a carefully designed upper bound on the AP-based loss function over weakly supervised samples. Using publicly available datasets, we demonstrate the advantage of our approach over standard loss-based binary classifiers on two challenging problems: action classification and character recognition.

Abstract

The accuracy of information retrieval systems is often measured using average precision (AP). Given a set of positive (relevant) and negative (non-relevant) samples, the parameters of a retrieval system can be estimated using the AP-SVM framework, which minimizes a regularized convex upper bound on the empirical AP loss. However, the high computational complexity of loss-augmented inference, which is required for learning an AP-SVM, prohibits its use with large training datasets. To alleviate this deficiency, we propose three complementary approaches. The first approach guarantees an asymptotic decrease in the computational complexity of loss-augmented inference by exploiting the problem structure. The second approach takes advantage of the fact that we do not require a full ranking during loss-augmented inference. This helps us to avoid the expensive step of sorting the negative samples according to their individual scores. The third approach approximates the AP loss over all samples by the AP loss over difficult samples (for example, those that are incorrectly classified by a binary SVM), while ensuring the correct classification of the remaining samples. Using the PASCAL VOC action classification and object detection datasets, we show that our approaches provide significant speed-ups during training without degrading the test accuracy of AP-SVM.

Abstract

In this paper, we present a practical and scalable retrieval framework for large-scale document image collections, for an Indian language script that does not have a robust OCR. OCR-based methods face difficulties in character segmentation and recognition, especially for the complex Indian language scripts. We realize that character recognition is only an intermediate step toward actually labeling words. Hence, we re-pose the problem as one of directly performing word annotation. This new approach has better recognition performance, as well as easier segmentation requirements. However, the number of classes in word annotation is much larger than those for character recognition, making such a classification scheme expensive to train and test. To address this issue, we present a novel framework that replaces naive classification with a carefully designed mixture of indexing and classification schemes. This enables us to build a search system over a large collection of 1,000 books of Telugu, consisting of 120K document images or 36M individual words. This is the largest searchable document image collection for a script without an OCR that we are aware of. Our retrieval system performs significantly well with a mean average precision of 0.8.

Abstract

This paper investigates the problem of cross document image retrieval, i.e. use of query images from one style (say font) to perform retrieval from a collection which is in a different style (say a different set of books). We present two approaches to tackle this problem. We propose an effective style independent retrieval scheme using a nonlinear style-content separation model. We also propose a semi-supervised style transfer strategy to expand the query into multiple styles. We validate both these approaches on a collection of word images which vary in fonts/styles