Cell segmentation is essential for obtaining quantitative data from biological microscopy images. Over the past decade, multiple supervised deep learning methods have been developed to aid in this process. However, these processes require extensive human-annotated labels to train the algorithms. To address this, we propose a novel pipeline for segmentation of images having higher cell-density. We implemented unsupervised DBSCAN to segment the low density (sparse cell) images. The segmented outputs from these sparse cell images were then used as labeled data to improve the segmentation of high-density (crowded cell) images using a supervised U-Net framework. The proposed pipeline has zero reliance on manual annotation and we achieved accurate segmentation of dense cells. This can be used for classification of images according to their packing density and decode the cell communication by computational modeling.

A Bus Rapid Transit System (BRTS) with earmarked lanes potentially provides efficient public transportation, and helps in controlling urban traffic congestion. While travel time prediction (TTP) is essential in a BRTS, existing algorithms generally assume GPS logs available at short uniform intervals. However, those are rarely evaluated on BRTS in emerging economies, where logged GPS data could be available at sparse nonuniform intervals. To fill the gap, we study the efficacy of certain well known ML models, namely, Random Forests (RF), Light Gradient Boosting (LGB), and Extreme Gradient Boosting (XGBoost, XGB) in utilizing historical data. Performance of those ensemble learning methods is compared with that of a conventional travel time prediction (CTTP) method, which uses historical averaging. It was found that XGB was superior to other methods at hand, and the prediction error by approximately 60% compared to the CTTP method. Alongside improving the experience of commuters, the proposed XGB-based TTP method also improves the estimation of intersection crossing time (ICT), which potentially leads to efficient traffic policy making.

Due to recent advent of light field visualization, ac-quisition/creation, encoding, transmission, rendering and quality assessment of 3D light field content has gained momentum. In particular, large light field displays need content with large field of view, and with high spatial and angular quality. Accordingly, subjective and objective quality evaluation studies have been conducted to examine spatial, angular and spatio-angular aspects of light field visualization. Recently, the effect of various zooming levels of the displayed content, as well as regions of interest on Quality of Experience (QoE) has also been explored. However, there has been no systematic attempt to see how the features of the content itself affect the visualization quality. In this work, we attempt to examine the effects of some primitive features of the content on subjective QoE. The results are based on a subjective study conducted on a large light field display, offering virtually continuous horizontal parallax.

In this paper, we present a high-angular-resolution data-set created using a turntable arrangement. Seven distinct objects, positioned on an automated turntable, were captured from three camera positions for every half degree of rotation, generating 720 images for each camera position. For each object, the camera positions were registered to the coordinate system of the middle camera. Intrinsic parameters of the camera were also estimated. A data-set of this kind is instrumental for research in a variety of areas, such as light field visualization, manifold learning, visual quality assessment, evaluation of preferred object orientation etc. Due to the availability of three-view stereo, this data-set could be useful for studying view interpolation techniques.

Air pollution poses disproportionate health hazard in developing countries, due to juxtaposition of industrial units and residences. While excessive emission is routinely detected, enforcement of emission norms remains rare due to present technological limitations in pinpointing sources. To fill this gap, we propose a method of source localization and emission rate estimation via AERMOD-based simulation under mean squared error criterion. Further, we validate the method using field data collected from an industrial cluster in India.

Wireless transfer of energy through directed radio frequency waves has the potential to realize perennially operating sensor nodes by replenishing the energy contained in the limited on-board battery. However, the high power energy transfer from energy transmitters (ETs) interferes with data communication, limiting the coexistence of these functions. This paper provides the first experimental study to quantify the rate of charging, packet loss due to interference, and suitable ranges for charging and data communication of the ETs. It also explores how the placement and relative distances of multiple ETs affect the charging process, demonstrating constructive and destructive energy aggregation at the sensor nodes. Finally, we investigate the impact of the separation in frequency between data and energy transmissions, as well as among multiple concurrent energy transmissions. Our results aim at providing insights on radio frequency-based energy harvesting wireless sensor networks for enhanced protocol design and network planning.

Can reliable telecardiology be achieved at low bandwidth cost? In response, we propose a detector at the user end so that only beats found to be anomalous are transmitted to a diagnostic center, where all received beats are correctly (re)classified. In this framework, high reliability is achieved by detectors with high sensitivity. Having laid the design framework, we then realize desired high-sensitivity detection using a dictionary learning approach. Specifically, using patient records from the MIT-BIH arrhythmia database, we detect ventricular ectopic beats (VEBs), which are known to be precursors to various serious arrhythmic conditions in the heart. In particular, we achieve a reliability of one undetected VEB in one thousand while saving 78.2% bandwidth using dictionaries with 240 atoms. With larger dictionaries with 420 atoms, we achieve an even higher bandwidth savings of 79.2% while allowing no (less than one in 1766) undetected VEB. Finally, we compare our results with performances a large set of reported heartbeat classifiers, and demonstrate the suitability of our approach in the context of telecardiology.

Harvesting energy from radio frequency (RF) waves brings us closer to achieving the goal for perpetual operation of a wireless sensor network (WSN) by replenishing the batteries of the sensor nodes. However, due to restrictions on the maximum transmitted power, path loss, and receiver sensitivity, only a small amount of energy can be harvested. While a dedicated RF source alleviates the problem to some extent, novel techniques are required to boost the energy transfer efficiency of the source. In this paper, we provide the first experimental demonstration of multi-path energy routing (MPER) for the case of a sparsely distributed WSNs and show its improved performance over direct energy transfer (DET). In addition, we extend this concept to the case of densely distributed WSNs and experimentally demonstrate and compare the gains obtained by 2- and 3-path energy routing over DET. Our experimental results show that significant energy gains can be achieved in a dense network deployment even when the node to be charged is partially blocked by the neighboring nodes.

Telecardiology systems, involving remote diagnosis of cardiac anomaly based on ECG signals, generally acquire such signals at the Nyquist rate, and transmits the data to diagnostic facilities. Such systems are not designed under either power or bandwidth constraints. However, in certain scenarios involving remote communities in developing and underdeveloped world, both the above constraints could be acute. The present paper takes a first step towards a constrained design keeping such scenarios in view. Specifically, we propose a system where automated classification is performed on the ECG signals, and only anomalous signals are transmitted for further diagnosis and intervention, thereby saving bandwidth. Additionally, we propose compressive sampling as a low-power alternative to traditional Nyquist sampling method, which also lowers bandwidth requirement. Finally, we illustrate our method by designing such a compressive classifier using ECG signals from the widely used PhysioNet database. Specifically, we demonstrate that an average down sampling factor of three leads to desirable classification performance in terms of both sensitivity and specificity while substantially saving both power and bandwidth.

Batteries of field nodes in a wireless sensor network pose an upper limit on the network lifetime. Energy harvesting and harvesting aware medium access control protocols have the potential to provide uninterrupted network operation, as they aim to replenish the lost energy so that energy neutral operation of the energy harvesting nodes can be achieved. To further improve the energy harvesting process, there is a need for novel schemes so that maximum energy is harvested in a minimum possible time. Multi-hop radio frequency (RF) energy transfer is one such solution that addresses these needs. With the optimal placement of energy relay nodes, multi-hop RF energy transfer can save energy of the source as well as time for the harvesting process. In this work we experimentally demonstrate multi-hop RF energy transfer, wherein two-hop energy transfer is shown to achieve significant energy and time savings with respect to the single-hop case. It is also shown that the gain obtained can be translated to energy transfer range extension.

The Körner-Marton modulo two-sum problem is described as a variation of the lossless two-help-one source coding problem.

We derive new algorithms for approximating the rate regions for a family of source coding problems that includes lossy source coding, lossy source coding with uncoded side information at the receiver (the Wyner-Ziv problem), and an achievability bound for lossy source coding with coded side in formation at the receiver. The new algorithms generalize a recent approximation algorithm by Go and Effros from lossless to lossy coding. In each case, prior information theoretic descriptions of the desired regions are available but difficult to evaluate for example sources due to their dependence on auxiliary random variables. Our algorithm builds a linear program whose solution is no less than the desired lower bound and no greater than (1 + ∈) times that optimal value. These guarantees are met even when the optimal value is unknown. Here ∈ ≫ 0 is a parameter chosen by the user; the algorithmic complexity grows as O(∈−M) as e approaches 0, where M ≫ 4 is a constant that depends on the source coding problem and the alphabet sizes of the sources.

Principles behind lossless and lossy coding are usually considered related, yet distinct. In contrast, we show that the direct statements of the rate-distortion theorem and the lossless coding theorem are consequences of a common distortion-abstracted phenomenon. Significantly, we extend such distortion abstraction to a more general multiterminal framework, and derive a canonical direct theorem that subsumes known results. Further, we show that the converse holds if all but at most one encoded sources are perfectly reconstructed, and, thereby, not only generalize known results but settle open problems such as the single-helper problem. More generally, a canonical sequence of inner bounds approaches the achievable region, and points to a new problem, whose solution, if found, would lead to a desirable computable description.

The issue of source/channel duality is as old as information theory itself. Whereas type covering has been identified as a basis for source coding and sphere packing as the basis for channel coding, very little effort has been directed at tying covering and packing together. In this paper, a covering result has been derived which gives asymptotically optimal packing as well. We propose our covering/packing lemma as a mathematical basis of duality. Our lemma also sheds light on outstanding issues of source coding such as explicit rate-distortion tradeoff and unifying basis for lossless and lossy coding

The Karhunen-Loeacuteve transform (KLT) is known to be optimal for high-rate transform coding of Gaussian vectors for both fixed-rate and variable-rate encoding. The KLT is also known to be suboptimal for some non-Gaussian models. This paper proves high-rate optimality of the KLT for variable-rate encoding of a broad class of non-Gaussian vectors: Gaussian vector-scale mixtures (GVSM), which extend the Gaussian scale mixture (GSM) model of natural signals. A key concavity property of the scalar GSM (same as the scalar GVSM) is derived to complete the proof. Optimality holds under a broad class of quadratic criteria, which include mean-squared error (MSE) as well as generalized f-divergence loss in estimation and binary classification systems. Finally, the theory is illustrated using two applications: signal estimation in multiplicative noise and joint optimization of classification/reconstruction systems

In a variety of applications, classification systems operate on compressed signals. The design of optimal transform coders optimizing a joint classification/reconstruction criterion was explored, where classification accuracy is measured using the Chernoff bound on probability of misclassification and reconstruction quality is measured using mean-squared error (MSE) distortion. Under a high-rate assumption, local optimality properties of the Karhunen-Loeve transform (KLT) for a certain class of Gaussian mixtures under a joint classification/MSE measure was shown. Analytical expressions for optimal bit-allocation were derived. This generalizes classical optimality properties of the KLT for Gaussian sources under the MSE criterion.

Mapping audio data to feature vectors for the classification, retrieval or identification tasks presents four principal challenges. The dimensionality of the input must be significantly reduced; the resulting features must be robust to likely distortions of the input; the features must be informative for the task at hand; and the feature extraction operation must be computationally efficient. We propose distortion discriminant analysis (DDA), which fulfills all four of these requirements. DDA constructs a linear, convolutional neural network out of layers, each of which performs an oriented PCA dimensional reduction. We demonstrate the effectiveness of DDA on two audio fingerprinting tasks: searching for 500 audio clips in 36 h of audio test data; and playing over 10 days of audio against a database with approximately 240 000 fingerprints. We show that the system is robust to kinds of noise that are not present in the training procedure. In the large test, the system gives a false positive rate of 1.5 /spl times/ 10/sup -8/ per audio clip, per fingerprint, at a false negative rate of 0.2% per clip.

A key problem faced by audio identification, classification, and retrieval systems is the mapping of high-dimensional audio input data into informative lower-dimensional feature vectors. This paper explores an automatic dimensionality reduction algorithm called Distortion Discriminant Analysis (DDA). Each layer of DDA projects its input into directions which maximize the SNR for a given set of distortions. Multiple layers efficiently extract features over a wide temporal window. The audio input to DDA undergoes perceptually-relevant preprocessing and de-equalization, to further suppress distortions. We apply DDA to the task of identifying audio clips in an incoming audio stream, based on matching stored audio fingerprints. We show excellent test results on matching input fingerprints against 36 hours of stored audio data.

In a variety of applications (including automatic target recognition) image classification algorithms operate on compressed image data. This paper explores the design of optimal transform coders and scalar quantizers using Chernoff bounds on probability of misclassification as the measure of classification accuracy. This design improves the classification performance but the mean square error (as well as the visual quality) of the coded image degrades. However, by appropriately combining classification accuracy and mean square error in the cost function, one can achieve good classification with low (visual) distortion, which is desirable in classification systems requiring visual authentication.