Sentence is a basic linguistic unit, however, little is known about how information content is distributed across different positions of a sentence. Based on authentic language data of English, the present study calculated the entropy and other entropy-related statistics for different sentence positions. The statistics indicate a three-step staircase-shaped distribution pattern, with entropy in the initial position lower than the medial positions (positions other than the initial and final), the medial positions lower than the final position and the medial positions showing no significant difference. The results suggest that: (1) the hypotheses of Constant Entropy Rate and Uniform Information Density do not hold for the sentence-medial positions; (2) the context of a word in a sentence should not be simply defined as all the words preceding it in the same sentence; and (3) the contextual information content in a sentence does not accumulate incrementally but follows a pattern of "the whole is greater than the sum of parts".
Understanding expressions of emotions in support forums has considerable value and NLP methods are key to automating this. Many approaches understandably use subjective categories which are more fine-grained than a straightforward polarity-based spectrum. However, the definition of such categories is non-trivial and, in fact, we argue for a need to incorporate communicative elements even beyond subjectivity. To support our position, we report experiments on a sentiment-labelled corpus of posts taken from a medical support forum. We argue that not only is a more fine-grained approach to text analysis important, but simultaneously recognising the social function behind affective expressions enable a more accurate and valuable level of understanding.
We describe the solution of team ISMLL for the ECML-PKDD 2016 Discovery Challenge on Bank Card Usage for both tasks. Our solution is based on three pillars. Gradient boosted decision trees as a strong regression and classification model, an intensive search for good hyperparameter configurations and strong features that exploit geolocation information. This approach achieved the best performance on the public leaderboard for the first task and a decent fourth position for the second task.
We describe a general method of detecting valid chains or links of pieces on a two-dimensional grid. Specifically, using the example of the chess variant known as Switch-Side Chain-Chess (SSCC). Presently, no foolproof method of detecting such chains in any given chess position is known and existing graph theory, to our knowledge, is unable to fully address this problem either. We therefore propose a solution implemented and tested using the C++ programming language. We have been unable to find an incorrect result and therefore offer it as the most viable solution thus far to the chain-detection problem in this chess variant. The algorithm is also scalable, in principle, to areas beyond two-dimensional grids such as 3D analysis and molecular chemistry.
Using neural networks to generate replies in human-computer dialogue systems is attracting increasing attention over the past few years. However, the performance is not satisfactory: the neural network tends to generate safe, universally relevant replies which carry little meaning. In this paper, we propose a content-introducing approach to neural network-based generative dialogue systems. We first use pointwise mutual information (PMI) to predict a noun as a keyword, reflecting the main gist of the reply. We then propose seq2BF, a "sequence to backward and forward sequences" model, which generates a reply containing the given keyword. Experimental results show that our approach significantly outperforms traditional sequence-to-sequence models in terms of human evaluation and the entropy measure, and that the predicted keyword can appear at an appropriate position in the reply.
In this paper, we introduce two indoor Wireless Local Area Network (WLAN) positioning methods using augmented sparse recovery algorithms. These schemes render a sparse user's position vector, and in parallel, minimize the distance between the online measurement and radio map. The overall localization scheme for both methods consists of three steps: 1) coarse localization, obtained from comparing the online measurements with clustered radio map. A novel graph-based method is proposed to cluster the offline fingerprints. In the online phase, a Region Of Interest (ROI) is selected within which we search for the user's location; 2) Access Point (AP) selection; and 3) fine localization through the novel sparse recovery algorithms. Since the online measurements are subject to inordinate measurement readings, called outliers, the sparse recovery methods are modified in order to jointly estimate the outliers and user's position vector. The outlier detection procedure identifies the APs whose readings are either not available or erroneous. The proposed localization methods have been tested with Received Signal Strength (RSS) measurements in a typical office environment and the results show that they can localize the user with significantly high accuracy and resolution which is superior to the results from competing WLAN fingerprinting localization methods.
Reordering is a preprocessing stage for Statistical Machine Translation (SMT) system where the words of the source sentence are reordered as per the syntax of the target language. We are proposing a rich set of rules for better reordering. The idea is to facilitate the training process by better alignments and parallel phrase extraction for a phrase-based SMT system. Reordering also helps the decoding process and hence improving the machine translation quality. We have observed significant improvements in the translation quality by using our approach over the baseline SMT. We have used BLEU, NIST, multi-reference word error rate, multi-reference position independent error rate for judging the improvements. We have exploited open source SMT toolkit MOSES to develop the system.
The basis for most vision based applications like robotics, self-driving cars and potentially augmented and virtual reality is a robust, continuous estimation of the position and orientation of a camera system w.r.t the observed environment (scene). In recent years many vision based systems that perform simultaneous localization and mapping (SLAM) have been presented and released as open source. In this paper, we extend and improve upon a state-of-the-art SLAM to make it applicable to arbitrary, rigidly coupled multi-camera systems (MCS) using the MultiCol model. In addition, we include a performance evaluation on accurate ground truth and compare the robustness of the proposed method to a single camera version of the SLAM system. An open source implementation of the proposed multi-fisheye camera SLAM system can be found on-line https://github.com/urbste/MultiCol-SLAM.
This position paper advocates a communications-inspired approach to the design of machine learning systems on energy-constrained embedded `always-on' platforms. The communications-inspired approach has two versions - 1) a deterministic version where existing low-power communication IC design methods are repurposed, and 2) a stochastic version referred to as Shannon-inspired statistical information processing employing information-based metrics, statistical error compensation (SEC), and retraining-based methods to implement ML systems on stochastic circuit/device fabrics operating at the limits of energy-efficiency. The communications-inspired approach has the potential to fully leverage the opportunities afforded by ML algorithms and applications in order to address the challenges inherent in their deployment on energy-constrained platforms.
Many online communities present user-contributed responses such as reviews of products and answers to questions. User-provided helpfulness votes can highlight the most useful responses, but voting is a social process that can gain momentum based on the popularity of responses and the polarity of existing votes. We propose the Chinese Voting Process (CVP) which models the evolution of helpfulness votes as a self-reinforcing process dependent on position and presentation biases. We evaluate this model on Amazon product reviews and more than 80 StackExchange forums, measuring the intrinsic quality of individual responses and behavioral coefficients of different communities.
Temporal common sense has applications in AI tasks such as QA, multi-document summarization, and human-AI communication. We propose the task of sequencing -- given a jumbled set of aligned image-caption pairs that belong to a story, the task is to sort them such that the output sequence forms a coherent story. We present multiple approaches, via unary (position) and pairwise (order) predictions, and their ensemble-based combinations, achieving strong results on this task. We use both text-based and image-based features, which depict complementary improvements. Using qualitative examples, we demonstrate that our models have learnt interesting aspects of temporal common sense.
OCR character segmentation for multilingual printed documents is difficult due to the diversity of different linguistic characters. Previous approaches mainly focus on monolingual texts and are not suitable for multilingual-lingual cases. In this work, we particularly tackle the Chinese/English mixed case by reframing it as a semantic segmentation problem. We take advantage of the successful architecture called fully convolutional networks (FCN) in the field of semantic segmentation. Given a wide enough receptive field, FCN can utilize the necessary context around a horizontal position to determinate whether this is a splitting point or not. As a deep neural architecture, FCN can automatically learn useful features from raw text line images. Although trained on synthesized samples with simulated random disturbance, our FCN model generalizes well to real-world samples. The experimental results show that our model significantly outperforms the previous methods.
This paper addresses the problem of 3D human pose estimation from a single image. We follow a standard two-step pipeline by first detecting the 2D position of the $N$ body joints, and then using these observations to infer 3D pose. For the first step, we use a recent CNN-based detector. For the second step, most existing approaches perform 2$N$-to-3$N$ regression of the Cartesian joint coordinates. We show that more precise pose estimates can be obtained by representing both the 2D and 3D human poses using $N\times N$ distance matrices, and formulating the problem as a 2D-to-3D distance matrix regression. For learning such a regressor we leverage on simple Neural Network architectures, which by construction, enforce positivity and symmetry of the predicted matrices. The approach has also the advantage to naturally handle missing observations and allowing to hypothesize the position of non-observed joints. Quantitative results on Humaneva and Human3.6M datasets demonstrate consistent performance gains over state-of-the-art. Qualitative evaluation on the images in-the-wild of the LSP dataset, using the regressor learned on Human3.6M, reveals very promising generalization results.
Tracking the movement of human eyes is expected to yield natural and convenient applications based on human-computer interaction (HCI). To implement an effective eye-tracking system, eye movements must be recorded without placing any restriction on the user's behavior or user discomfort. This paper describes an eye movement recording system that offers free-head, simple configuration. It does not require the user to wear anything on her head, and she can move her head freely. Instead of using a computer, the system uses a visual digital signal processor (DSP) camera to detect the position of eye corner, the center of pupil and then calculate the eye movement. Evaluation tests show that the sampling rate of the system can be 300 Hz and the accuracy is about 1.8 degree/s.
Combining new, low-cost thermal infrared and time-of-flight range sensors provides new opportunities. In this position paper we explore the possibilities of combining these sensors and using their fused data for person detection. The proposed calibration approach for this sensor combination differs from the traditional stereo camera calibration in two fundamental ways. A first distinction is that the spectral sensitivity of the two sensors differs significantly. In fact, there is no sensitivity range overlap at all. A second distinction is that their resolution is typically very low, which requires special attention. We assume a situation in which the sensors' relative position is known, but their orientation is unknown. In addition, some of the typical measurement errors are discussed, and methods to compensate for them are proposed. We discuss how the fused data could allow increased accuracy and robustness without the need for complex algorithms requiring large amounts of computational power and training data.
Scarce data is a major challenge to scaling robot learning to truly complex tasks, as we need to generalize locally learned policies over different "contexts". Bayesian optimization approaches to contextual policy search (CPS) offer data-efficient policy learning that generalize over a context space. We propose to improve data-efficiency by factoring typically considered contexts into two components: target-type contexts that correspond to a desired outcome of the learned behavior, e.g. target position for throwing a ball; and environment type contexts that correspond to some state of the environment, e.g. initial ball position or wind speed. Our key observation is that experience can be directly generalized over target-type contexts. Based on that we introduce Factored Contextual Policy Search with Bayesian Optimization for both passive and active learning settings. Preliminary results show faster policy generalization on a simulated toy problem. A full paper extension is available at arXiv:1904.11761
Conventional approaches to image de-fencing suffer from non-robust fence detection and are limited to processing images of static scenes. In this position paper, we propose an automatic de-fencing algorithm for images of dynamic scenes. We divide the problem of image de-fencing into the tasks of automated fence detection, motion estimation and fusion of data from multiple frames of a captured video of the dynamic scene. Fences are detected automatically using two approaches, namely, employing Gabor filter and a machine learning method. We cast the fence removal problem in an optimization framework, by modeling the formation of the degraded observations. The inverse problem is solved using split Bregman technique assuming total variation of the de-fenced image as the regularization constraint.
We propose a method that integrates two widely available data sources, building footprints from 2D maps and street level images, to derive valuable information that is generally difficult to acquire -- building heights and building facade masks in images. Building footprints are elevated in world coordinates and projected onto images. Building heights are estimated by scoring projected footprints based on their alignment with building features in images. Building footprints with estimated heights can be converted to simple 3D building models, which are projected back to images to identify buildings. In this procedure, accurate camera projections are critical. However, camera position errors inherited from external sensors commonly exist, which adversely affect results. We derive a solution to precisely locate cameras on maps using correspondence between image features and building footprints. Experiments on real-world datasets show the promise of our method.
This paper addresses the matching of short music audio snippets to the corresponding pixel location in images of sheet music. A system is presented that simultaneously learns to read notes, listens to music and matches the currently played music to its corresponding notes in the sheet. It consists of an end-to-end multi-modal convolutional neural network that takes as input images of sheet music and spectrograms of the respective audio snippets. It learns to predict, for a given unseen audio snippet (covering approximately one bar of music), the corresponding position in the respective score line. Our results suggest that with the use of (deep) neural networks -- which have proven to be powerful image processing models -- working with sheet music becomes feasible and a promising future research direction.
We present a novel, automatic eye gaze tracking scheme inspired by smooth pursuit eye motion while playing mobile games or watching virtual reality contents. Our algorithm continuously calibrates an eye tracking system for a head mounted display. This eliminates the need for an explicit calibration step and automatically compensates for small movements of the headset with respect to the head. The algorithm finds correspondences between corneal motion and screen space motion, and uses these to generate Gaussian Process Regression models. A combination of those models provides a continuous mapping from corneal position to screen space position. Accuracy is nearly as good as achieved with an explicit calibration step.