Search

Realizing deterministic, high-fidelity entangling interactions--of the kind that can be utilized for efficient quantum information processing--between photons remains an elusive goal. Here, we address this long-standing issue by devising a protocol for creating and manipulating highly-entangled superpositions of well-controlled states of light by using an on-chip photonic system that has recently been shown to implement three-dimensional, non-Abelian quantum holonomy. Our calculations indicate that a subset of such entangled superpositions are maximally-entangled, volume-law states, and that the underlying entanglement can be distilled and purified for applications in quantum science. Specifically, we envisage that this entangling mechanism could be utilized for realizing two-qubit, universal, entangling quantum gates with linear photonic elements alone. Furthermore, this method should also--at least, in principle--be generalizable to systems of neutral atoms., Comment: 7 pages including references, 5 figures; no accompanying Supplemental Material

Analog quantum simulation - the technique of using one experimentally well-controlled physical system to mimic the behavior of another - has quickly emerged as one of the most promising near term strategies for studying strongly correlated quantum many-body systems. In particular, systems of interacting photons, realizable in solid-state cavity and circuit QED frameworks, for example, hold tremendous promise for the study of nonequilibrium many-body phenomena due to the capability to locally create and destroy photons. These systems are typically modeled using a Jaynes-Cummings-Hubbard (JCH) Hamiltonian, named due to similarities with the Bose-Hubbard (BH) model. Here, we present a non-perturbative procedure for transforming the JC Hamiltonian into a dressed operator representation that, in its most general form, admits an infinite sum of bosonic $k$-body terms where $k$ is bound only by the number of excitations in the system. We closely examine this result in both the dispersive and resonant coupling regimes, finding rapid convergence in the former and contributions from $k\gg1$ in the latter. Through extension to a two-site JCH system, we demonstrate that this approach facilitates close inspection of the analogy between the JCH and BH models and its breakdown for resonant light-matter coupling. Finally, we use this framework to survey the many-body character of a two-site JCH for general system parameters, identifying four unique quantum phases and the parameter regimes in which they are realized, thus highlighting phenomena realizable with finite JCH-based quantum simulators beyond the BH model. More broadly, this work is intended to serve as a clear mathematical exposition of bosonic many-body interactions underlying JC-type systems, often postulated through analogy to Kerr-like nonlinear susceptibilities or by matching coefficients to obtain the appropriate eigenvalue spectrum., Comment: 25 pages, 6 figures

We propose MoodNet - A Deep Convolutional Neural Network based architecture to effectively predict the emotion associated with a piece of music given its audio and lyrical content.We evaluate different architectures consisting of varying number of two-dimensional convolutional and subsampling layers,followed by dense layers.We use Mel-Spectrograms to represent the audio content and word embeddings-specifically 100 dimensional word vectors, to represent the textual content represented by the lyrics.We feed input data from both modalities to our MoodNet architecture.The output from both the modalities are then fused as a fully connected layer and softmax classfier is used to predict the category of emotion.Using F1-score as our metric,our results show excellent performance of MoodNet over the two datasets we experimented on-The MIREX Multimodal dataset and the Million Song Dataset.Our experiments reflect the hypothesis that more complex models perform better with more training data.We also observe that lyrics outperform audio as a better expressed modality and conclude that combining and using features from multiple modalities for prediction tasks result in superior performance in comparison to using a single modality as input., Comment: 6 pages

We report the optical excitation-dependent output photocurrent characteristics of GaN-based polariton diode lasers operated under reverse-bias at room temperature. The photocurrent demonstrates a non-linear enhancement at an incident optical power of ~ 1.6 mW, which is approximately equivalent to the value of polariton lasing threshold observed when the diodes are operated under forward bias conditions. This is explained in the framework of an Auger-like process of excitonic dissociation into its constituent electron-hole pairs, which can be stimulated by the occupation of the polariton lasing states. The observed effect is a remarkable manifestation of the bosonic final state stimulation in polariton lasers. A model based on the coupled kinetic equations for the free carriers, the excitonic reservoir and the polariton condensate shows a good agreement to the experimental data.

Automation of brain matter segmentation from MR images is a challenging task due to the irregular boundaries between the grey and white matter regions. In addition, the presence of intensity inhomogeneity in the MR images further complicates the problem. In this paper, we propose a texture and vesselness incorporated version of the ScatterNet Hybrid Deep Learning Network (TS-SHDL) that extracts hierarchical invariant mid-level features, used by fisher vector encoding and a conditional random field (CRF) to perform the desired segmentation. The performance of the proposed network is evaluated by extensive experimentation and comparison with the state-of-the-art methods on several 2D MRI scans taken from the synthetic McGill Brain Web as well as on the MRBrainS dataset of real 3D MRI scans. The advantages of the TS-SHDL network over supervised deep learning networks is also presented in addition to its superior performance over the state-of-the-art., Comment: To Appear in the IEEE International Conference on Computer Vision Workshops (ICCVW) 2017