Search

State-of-the-art medical multi-modal large language models (med-MLLM), like LLaVA-Med or BioMedGPT, leverage instruction-following data in pre-training. However, those models primarily focus on scaling the model size and data volume to boost performance while mainly relying on the autoregressive learning objectives. Surprisingly, we reveal that such learning schemes might result in a weak alignment between vision and language modalities, making these models highly reliant on extensive pre-training datasets - a significant challenge in medical domains due to the expensive and time-consuming nature of curating high-quality instruction-following instances. We address this with LoGra-Med, a new multi-graph alignment algorithm that enforces triplet correlations across image modalities, conversation-based descriptions, and extended captions. This helps the model capture contextual meaning, handle linguistic variability, and build cross-modal associations between visuals and text. To scale our approach, we designed an efficient end-to-end learning scheme using black-box gradient estimation, enabling faster LLaMa 7B training. Our results show LoGra-Med matches LLAVA-Med performance on 600K image-text pairs for Medical VQA and significantly outperforms it when trained on 10% of the data. For example, on VQA-RAD, we exceed LLAVA-Med by 20.13% and nearly match the 100% pre-training score (72.52% vs. 72.64%). We also surpass SOTA methods like BiomedGPT on visual chatbots and RadFM on zero-shot image classification with VQA, highlighting the effectiveness of multi-graph alignment., Comment: First version

The Coulomb impurity problem of graphene, in the absence of a magnetic field, displays discrete scale invariance. Applying a magnetic field introduces a new magnetic length scale $\ell$ and breaks discrete scale invariance. Moreover, a magnetic field is a singular perturbation as it turns complex energies into real energies. Nonetheless, the Coulomb potential must be regularized with a length $R$ at short distances for supercritical impurities. We investigate the structure of the induced density of a filled Landau impurity band in the supercritical regime. The coupling between Landau level states by the impurity potential is nontrivial and can lead to several anomalous effects. First, we find that the peak in the induced density can be located away from the center of the impurity, depending on the characteristics of the Landau impurity bands. Second, the impurity charge is screened, despite the Landau impurity band being filled. Third, anticrossing impurity states lead to additional impurity cyclotron resonances.

The realization of electron-spin resonance at the single-atom level using scanning tunneling microscopy has opened new avenues for coherent quantum sensing and quantum state manipulation at the ultimate size limit. This allows to build many-body Hamiltonians and the study of their complex physical behavior. Recently, a novel qubit platform has emerged from this field, raising questions about the driving mechanism from single-atom magnets. In this work, we demonstrate how single-atom magnets can be used to drive a nearby single spin qubit efficiently, while also addressing critical aspects related to the optimization of experimental parameters.

Bundle recommendation aims to enhance business profitability and user convenience by suggesting a set of interconnected items. In real-world scenarios, leveraging the impact of asymmetric item affiliations is crucial for effective bundle modeling and understanding user preferences. To address this, we present BunCa, a novel bundle recommendation approach employing item-level causation-enhanced multi-view learning. BunCa provides comprehensive representations of users and bundles through two views: the Coherent View, leveraging the Multi-Prospect Causation Network for causation-sensitive relations among items, and the Cohesive View, employing LightGCN for information propagation among users and bundles. Modeling user preferences and bundle construction combined from both views ensures rigorous cohesion in direct user-bundle interactions through the Cohesive View and captures explicit intents through the Coherent View. Simultaneously, the integration of concrete and discrete contrastive learning optimizes the consistency and self-discrimination of multi-view representations. Extensive experiments with BunCa on three benchmark datasets demonstrate the effectiveness of this novel research and validate our hypothesis.

We study a new variant of graph coloring by adding a connectivity constraint. A path in a vertex-colored graph is called conflict-free if there is a color that appears exactly once on its vertices. A connected graph $G$ is said to be strongly conflict-free vertex-connection $k$-colorable if $G$ admits a vertex $k$-coloring such that any two distinct vertices of $G$ are connected by a conflict-free $shortest$ path. Among others, we show that deciding whether a given graph is strongly conflict-free vertex-connection $3$-colorable is NP-complete even when restricted to $3$-colorable graphs with diameter $3$, radius $2$ and domination number $3$, and, assuming the Exponential Time Hypothesis (ETH), cannot be solved in $2^{o(n)}$ time on such restricted input graphs with $n$ vertices. This hardness result is quite strong when compared to the ordinary $3$-COLORING problem: it is known that $3$-COLORING is solvable in polynomial time in graphs with bounded domination number, and assuming ETH, cannot be solved in $2^{o(\sqrt{n})}$ time in $n$-vertex graphs with diameter $3$ and radius $2$. On the positive side, we point out that a strong conflict-free vertex-connection coloring with minimum color number of a given split graph or a co-bipartite graph can be computed in polynomial time., Comment: The full version of a COCOON 2024 paper

This paper proposes a joint design of probabilistic constellation shaping (PCS) and precoding to enhance the sum-rate performance of multi-user visible light communications (VLC) broadcast channels subject to signal amplitude constraint. In the proposed design, the transmission probabilities of bipolar $M$-pulse amplitude modulation ($M$-PAM) symbols for each user and the transmit precoding matrix are jointly optimized to improve the sum-rate performance. The joint design problem is shown to be a complex non-convex problem due to the non-convexity of the objective function. To tackle the problem, the firefly algorithm (FA), a nature-inspired heuristic optimization approach, is employed to solve a local optima to the original non-convex optimization problem. The FA-based approach, however, suffers from high computational complexity. Therefore, we propose a low-complexity design based on zero-forcing (ZF) precoding, which is solved using an alternating optimization (AO) approach. Simulation results reveal that the proposed joint design with PCS significantly improves the sum-rate performance compared to the conventional design with uniform signaling. Some insights into the optimal symbol distributions of the two joint design approaches are also provided.

This study investigates the characteristic polarization formation and evolution of vector dark matter (VDM) in the outer halo of galaxies. By employing numerical simulations, we analyze the behavior of VDM under different initial conditions -- homogeneous, isotropic, and partially polarized. The simulations solve the Schr\"odinger-Poisson equations, examining the spin density distribution and its evolution during gravitational collapse and halo formation. Our results reveal that VDM forms halos and central Proca stars from homogeneous and isotropic conditions, with the polarization density fluctuation amplitude mirroring VDM matter density. In scenarios with no initial polarization, spin density remains stable in the halo core but fluctuates in outer regions. Partially polarized initial conditions lead to a conservation of total polarization, with increased core polarization resulting in opposite polarization in the periphery. We examine the novel consequences of the partially polarized state for direct detection of dark photons, i.e., VDM kinetically mixed with ordinary photons., Comment: 12 pages,12 figures

Comprehending how humans process visual information in dynamic settings is crucial for psychology and designing user-centered interactions. While mobile eye-tracking systems combining egocentric video and gaze signals can offer valuable insights, manual analysis of these recordings is time-intensive. In this work, we present a novel human-centered learning algorithm designed for automated object recognition within mobile eye-tracking settings. Our approach seamlessly integrates an object detector with a spatial relation-aware inductive message-passing network (I-MPN), harnessing node profile information and capturing object correlations. Such mechanisms enable us to learn embedding functions capable of generalizing to new object angle views, facilitating rapid adaptation and efficient reasoning in dynamic contexts as users navigate their environment. Through experiments conducted on three distinct video sequences, our interactive-based method showcases significant performance improvements over fixed training/testing algorithms, even when trained on considerably smaller annotated samples collected through user feedback. Furthermore, we demonstrate exceptional efficiency in data annotation processes and surpass prior interactive methods that use complete object detectors, combine detectors with convolutional networks, or employ interactive video segmentation., Comment: Updated version

Partial multi-task learning where training examples are annotated for one of the target tasks is a promising idea in remote sensing as it allows combining datasets annotated for different tasks and predicting more tasks with fewer network parameters. The na\"ive approach to partial multi-task learning is sub-optimal due to the lack of all-task annotations for learning joint representations. This paper proposes using knowledge distillation to replace the need of ground truths for the alternate task and enhance the performance of such approach. Experiments conducted on the public ISPRS 2D Semantic Labeling Contest dataset show the effectiveness of the proposed idea on partial multi-task learning for semantic tasks including object detection and semantic segmentation in aerial images., Comment: Accepted for oral presentation at IGARSS 2024

Advances in rendering have led to tremendous growth in texture assets, including resolution, complexity, and novel textures components, but this growth in data volume has not been matched by advances in its compression. Meanwhile Neural Image Compression (NIC) has advanced significantly and shown promising results, but the proposed methods cannot be directly adapted to neural texture compression. First, texture compression requires on-demand and real-time decoding with random access during parallel rendering (e.g. block texture decompression on GPUs). Additionally, NIC does not support multi-resolution reconstruction (mip-levels), nor does it have the ability to efficiently jointly compress different sets of texture channels. In this work, we introduce a novel approach to texture set compression that integrates traditional GPU texture representation and NIC techniques, designed to enable random access and support many-channel texture sets. To achieve this goal, we propose an asymmetric auto-encoder framework that employs a convolutional encoder to capture detailed information in a bottleneck-latent space, and at decoder side we utilize a fully connected network, whose inputs are sampled latent features plus positional information, for a given texture coordinate and mip level. This latent data is defined to enable simplified access to multi-resolution data by simply changing the scanning strides. Experimental results demonstrate that this approach provides much better results than conventional texture compression, and significant improvement over the latest method using neural networks., Comment: ECCV submission

The rise of new video modalities like virtual reality or autonomous driving has increased the demand for efficient multi-view video compression methods, both in terms of rate-distortion (R-D) performance and in terms of delay and runtime. While most recent stereo video compression approaches have shown promising performance, they compress left and right views sequentially, leading to poor parallelization and runtime performance. This work presents Low-Latency neural codec for Stereo video Streaming (LLSS), a novel parallel stereo video coding method designed for fast and efficient low-latency stereo video streaming. Instead of using a sequential cross-view motion compensation like existing methods, LLSS introduces a bidirectional feature shifting module to directly exploit mutual information among views and encode them effectively with a joint cross-view prior model for entropy coding. Thanks to this design, LLSS processes left and right views in parallel, minimizing latency; all while substantially improving R-D performance compared to both existing neural and conventional codecs., Comment: Accepted by CVPR2024

In the realm of Federated Learning (FL) applied to remote sensing image classification, this study introduces and assesses several innovative communication strategies. Our exploration includes feature-centric communication, pseudo-weight amalgamation, and a combined method utilizing both weights and features. Experiments conducted on two public scene classification datasets unveil the effectiveness of these strategies, showcasing accelerated convergence, heightened privacy, and reduced network information exchange. This research provides valuable insights into the implications of feature-centric communication in FL, offering potential applications tailored for remote sensing scenarios., Comment: 5 pages, to appear in IGARSS 2024

Uncertainty in medical image segmentation tasks, especially inter-rater variability, arising from differences in interpretations and annotations by various experts, presents a significant challenge in achieving consistent and reliable image segmentation. This variability not only reflects the inherent complexity and subjective nature of medical image interpretation but also directly impacts the development and evaluation of automated segmentation algorithms. Accurately modeling and quantifying this variability is essential for enhancing the robustness and clinical applicability of these algorithms. We report the set-up and summarize the benchmark results of the Quantification of Uncertainties in Biomedical Image Quantification Challenge (QUBIQ), which was organized in conjunction with International Conferences on Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2020 and 2021. The challenge focuses on the uncertainty quantification of medical image segmentation which considers the omnipresence of inter-rater variability in imaging datasets. The large collection of images with multi-rater annotations features various modalities such as MRI and CT; various organs such as the brain, prostate, kidney, and pancreas; and different image dimensions 2D-vs-3D. A total of 24 teams submitted different solutions to the problem, combining various baseline models, Bayesian neural networks, and ensemble model techniques. The obtained results indicate the importance of the ensemble models, as well as the need for further research to develop efficient 3D methods for uncertainty quantification methods in 3D segmentation tasks., Comment: initial technical report

This paper presents a comprehensive overview of the Comparative Opinion Mining from Vietnamese Product Reviews shared task (ComOM), held as part of the 10$^{th}$ International Workshop on Vietnamese Language and Speech Processing (VLSP 2023). The primary objective of this shared task is to advance the field of natural language processing by developing techniques that proficiently extract comparative opinions from Vietnamese product reviews. Participants are challenged to propose models that adeptly extract a comparative "quintuple" from a comparative sentence, encompassing Subject, Object, Aspect, Predicate, and Comparison Type Label. We construct a human-annotated dataset comprising $120$ documents, encompassing $7427$ non-comparative sentences and $2468$ comparisons within $1798$ sentences. Participating models undergo evaluation and ranking based on the Exact match macro-averaged quintuple F1 score., Comment: In Proceedings of VLSP 2023

Real-world environments are inherently non-stationary, frequently introducing new classes over time. This is especially common in time series classification, such as the emergence of new disease classification in healthcare or the addition of new activities in human activity recognition. In such cases, a learning system is required to assimilate novel classes effectively while avoiding catastrophic forgetting of the old ones, which gives rise to the Class-incremental Learning (CIL) problem. However, despite the encouraging progress in the image and language domains, CIL for time series data remains relatively understudied. Existing studies suffer from inconsistent experimental designs, necessitating a comprehensive evaluation and benchmarking of methods across a wide range of datasets. To this end, we first present an overview of the Time Series Class-incremental Learning (TSCIL) problem, highlight its unique challenges, and cover the advanced methodologies. Further, based on standardized settings, we develop a unified experimental framework that supports the rapid development of new algorithms, easy integration of new datasets, and standardization of the evaluation process. Using this framework, we conduct a comprehensive evaluation of various generic and time-series-specific CIL methods in both standard and privacy-sensitive scenarios. Our extensive experiments not only provide a standard baseline to support future research but also shed light on the impact of various design factors such as normalization layers or memory budget thresholds. Codes are available at https://github.com/zqiao11/TSCIL., Comment: Accepted by SIGKDD 2024 (ADS track). Codes available at https://github.com/zqiao11/TSCIL

The well-known Cluster Vertex Deletion problem (CVD) asks for a given graph $G$ and an integer $k$ whether it is possible to delete a set $S$ of at most $k$ vertices of $G$ such that the resulting graph $G-S$ is a cluster graph (a disjoint union of cliques). We give a complete characterization of graphs $H$ for which CVD on $H$-free graphs is polynomially solvable and for which it is NP-complete. Moreover, in the NP-completeness cases, CVD cannot be solved in sub-exponential time in the vertex number of the $H$-free input graphs unless the Exponential-Time Hypothesis fails. We also consider the connected variant of CVD, the Connected Cluster Vertex Deletion problem (CCVD), in which the set $S$ has to induce a connected subgraph of $G$. It turns out that CCVD admits the same complexity dichotomy for $H$-free graphs. Our results enlarge a list of rare dichotomy theorems for well-studied problems on $H$-free graphs., Comment: Extended version of a MFCS 2022 paper. To appear in Theory of Computing Systems

This paper introduces the Orthogonal Polynomials Quadrature Algorithm for Survival Analysis (OPSurv), a new method providing time-continuous functional outputs for both single and competing risks scenarios in survival analysis. OPSurv utilizes the initial zero condition of the Cumulative Incidence function and a unique decomposition of probability densities using orthogonal polynomials, allowing it to learn functional approximation coefficients for each risk event and construct Cumulative Incidence Function estimates via Gauss--Legendre quadrature. This approach effectively counters overfitting, particularly in competing risks scenarios, enhancing model expressiveness and control. The paper further details empirical validations and theoretical justifications of OPSurv, highlighting its robust performance as an advancement in survival analysis with competing risks.

Video compression systems must support increasing bandwidth and data throughput at low cost and power, and can be limited by entropy coding bottlenecks. Efficiency can be greatly improved by parallelizing coding, which can be done at much larger scales with new neural-based codecs, but with some compression loss related to data organization. We analyze the bit rate overhead needed to support multiple bitstreams for concurrent decoding, and for its minimization propose a method for compressing parallel-decoding entry points, using bidirectional bitstream packing, and a new form of jointly optimizing arithmetic coding termination. It is shown that those techniques significantly lower the overhead, making it easier to reduce it to a small fraction of the average bitstream size, like, for example, less than 1% and 0.1% when the average number of bitstream bytes is respectively larger than 95 and 1,200 bytes.

Multi-task partially annotated data where each data point is annotated for only a single task are potentially helpful for data scarcity if a network can leverage the inter-task relationship. In this paper, we study the joint learning of object detection and semantic segmentation, the two most popular vision problems, from multi-task data with partial annotations. Extensive experiments are performed to evaluate each task performance and explore their complementarity when a multi-task network cannot optimize both tasks simultaneously. We propose employing knowledge distillation to leverage joint-task optimization. The experimental results show favorable results for multi-task learning and knowledge distillation over single-task learning and even full supervision scenario. All code and data splits are available at https://github.com/lhoangan/multas, Comment: Accepted for publishing at BMVC 2023

Graphene zigzag nanoribbons, initially in a topologically ordered state, undergo a topological phase transition into crossover phases distinguished by quasi-topological order. We computed mutual information for both the topologically ordered phase and its crossover phases, revealing the following results: (i) In the topologically ordered phase, A-chirality carbon lines strongly entangle with B-chirality carbon lines on the opposite side of the zigzag ribbon. This entanglement persists but weakens in crossover phases. (ii) The upper zigzag edge entangles with non-edge lines of different chirality on the opposite side of the ribbon. (iii) Entanglement increases as more carbon lines are grouped together, regardless of the lines' chirality. No long-range entanglement was found in the symmetry-protected phase in the absence of disorder., Comment: A featured paper published in a special issue titled 'Entanglement Entropy and Quantum Phase Transitions' within the journal 'Entropy'. (This is the published version.)

Neural video codecs have recently become competitive with standard codecs such as HEVC in the low-delay setting. However, most neural codecs are large floating-point networks that use pixel-dense warping operations for temporal modeling, making them too computationally expensive for deployment on mobile devices. Recent work has demonstrated that running a neural decoder in real time on mobile is feasible, but shows this only for 720p RGB video. This work presents the first neural video codec that decodes 1080p YUV420 video in real time on a mobile device. Our codec relies on two major contributions. First, we design an efficient codec that uses a block-based motion compensation algorithm available on the warping core of the mobile accelerator, and we show how to quantize this model to integer precision. Second, we implement a fast decoder pipeline that concurrently runs neural network components on the neural signal processor, parallel entropy coding on the mobile GPU, and warping on the warping core. Our codec outperforms the previous on-device codec by a large margin with up to 48% BD-rate savings, while reducing the MAC count on the receiver side by $10 \times$. We perform a careful ablation to demonstrate the effect of the introduced motion compensation scheme, and ablate the effect of model quantization., Comment: Matches version published at WACV 2024

We present a semi-microscopic study of the 16C(d,p)17C transfer reaction. The 17C overlap integrals and spectroscopic factors are obtained from a microscopic cluster model, involving many 16C+n configurations. This microscopic model provides a fair description of the 17C bound-state energies. The 16C+d scattering wave functions are defined in the CDCC method, where the deuteron breakup is simulated by pseudostates. The transfer cross sections are in good agreement with recent data. We confirm the 16C(2+)+n structure of the ground state, and show that deuteron breakup effects have a significant influence on the cross sections. We study the 17C(p,d)16C reverse reaction and suggest that the cross section to the 2+ state should be large. A measurement of the ground-state cross section would provide a strong test of the microscopic wave functions., Comment: 8 pages, 5 figures, accepted for publication in Phys. Rev. C

Self-training allows a network to learn from the predictions of a more complicated model, thus often requires well-trained teacher models and mixture of teacher-student data while multi-task learning jointly optimizes different targets to learn salient interrelationship and requires multi-task annotations for each training example. These frameworks, despite being particularly data demanding have potentials for data exploitation if such assumptions can be relaxed. In this paper, we compare self-training object detection under the deficiency of teacher training data where students are trained on unseen examples by the teacher, and multi-task learning with partially annotated data, i.e. single-task annotation per training example. Both scenarios have their own limitation but potentially helpful with limited annotated data. Experimental results show the improvement of performance when using a weak teacher with unseen data for training a multi-task student. Despite the limited setup we believe the experimental results show the potential of multi-task knowledge distillation and self-training, which could be beneficial for future study. Source code is at https://lhoangan.github.io/multas., Comment: Accepted for International Conference in Computer Vision workshop (ICCVW) 2023

A common approach to solve inverse imaging problems relies on finding a maximum a posteriori (MAP) estimate of the original unknown image, by solving a minimization problem. In thiscontext, iterative proximal algorithms are widely used, enabling to handle non-smooth functions and linear operators. Recently, these algorithms have been paired with deep learning strategies, to further improve the estimate quality. In particular, proximal neural networks (PNNs) have been introduced, obtained by unrolling a proximal algorithm as for finding a MAP estimate, but over a fixed number of iterations, with learned linear operators and parameters. As PNNs are based on optimization theory, they are very flexible, and can be adapted to any image restoration task, as soon as a proximal algorithm can solve it. They further have much lighter architectures than traditional networks. In this article we propose a unified framework to build PNNs for the Gaussian denoising task, based on both the dual-FB and the primal-dual Chambolle-Pock algorithms. We further show that accelerated inertial versions of these algorithms enable skip connections in the associated NN layers. We propose different learning strategies for our PNN framework, and investigate their robustness (Lipschitz property) and denoising efficiency. Finally, we assess the robustness of our PNNs when plugged in a forward-backward algorithm for an image deblurring problem.

Knowledge distillation, a well-known model compression technique, is an active research area in both computer vision and remote sensing communities. In this paper, we evaluate in a remote sensing context various off-the-shelf object detection knowledge distillation methods which have been originally developed on generic computer vision datasets such as Pascal VOC. In particular, methods covering both logit mimicking and feature imitation approaches are applied for vehicle detection using the well-known benchmarks such as xView and VEDAI datasets. Extensive experiments are performed to compare the relative performance and interrelationships of the methods. Experimental results show high variations and confirm the importance of result aggregation and cross validation on remote sensing datasets., Comment: Accepted for publishing at IGARSS 2023

In a graph, a (perfect) matching cut is an edge cut that is a (perfect) matching. Matching Cut (MC), respectively, Perfect Matching Cut (PMC), is the problem of deciding whether a given graph has a matching cut, respectively, a perfect matching cut. The Disconnected Perfect Matching problem (DPM) is to decide if a graph has a perfect matching that contains a matching cut. Solving an open problem posed in [Lucke, Paulusma, Ries (ISAAC 2022, Algorithmica 2023) & Feghali, Lucke, Paulusma, Ries (arXiv:2212.12317)], we show that PMC is NP-complete in graphs without induced $14$-vertex path $P_{14}$. Our reduction also works simultaneously for MC and DPM, improving the previous hardness results of MC on $P_{15}$-free graphs and of DPM on $P_{19}$-free graphs to $P_{14}$-free graphs for both problems. Actually, we prove a slightly stronger result: within $P_{14}$-free $8$-chordal graphs (graphs without chordless cycles of length at least $9$), it is hard to distinguish between those without matching cuts (respectively, perfect matching cuts, disconnected perfect matchings) and those in which every matching cut is a perfect matching cut. Moreover, assuming the Exponential Time Hypothesis, none of these problems can be solved in $2^{o(n)}$ time for $n$-vertex $P_{14}$-free $8$-chordal graphs. On the positive side, we show that, as for MC [Moshi (JGT 1989)], DPM and PMC are polynomially solvable when restricted to $4$-chordal graphs. Together with the negative results, this partly answers an open question on the complexity of PMC in $k$-chordal graphs asked in [Le, Telle (TCS 2022) & Lucke, Paulusma, Ries (MFCS 2023)]., Comment: Extended version of a WG 2023 paper

We computed the phase diagram of the zigzag graphene nanoribbons as a function of on-site repulsion, doping, and disorder strength. The topologically ordered phase undergoes topological phase transitions into crossover phases, which are new disordered phases with a nonuniversal topological entanglement entropy with significant variance. The topological order is destroyed by competition between localization effects and on-site repulsion. We found that strong on-site repulsion and/or doping weakens the nonlocal correlations between the opposite zigzag edges. In one of the crossover phases, both $\frac{e^-}{2}$ fractional charges and spin-charge separation were absent; however, charge-transfer correlations between the zigzag edges were possible. Another crossover phase contains $\frac{e^-}{2}$ fractional charges, but no charge transfer correlations. In low-doped zigzag ribbons the interplay between electron localization and on-site repulsion contributes to the spatial separation of quasi-degenerate gap-edge states and protects the charge fractionalization against quantum fluctuations. In all these effects, mixed chiral gap-edge states play an important role. The properties of nontopological strongly disordered and strongly repulsive phases are also observed. Each phase of the phase diagram has a different zigzag-edge structure. Additionally, we investigated the tunneling of solitonic fractional charges under an applied voltage between the zigzag edges of undoped topologically ordered zigzag ribbons, and found that it may lead to a zero-bias tunneling anomaly., Comment: Published version, 2024 J. Phys.: Condens. Matter 36 265604

This paper presents a simulation platform, namely CIMulator, for quantifying the efficacy of various synaptic devices in neuromorphic accelerators for different neural network architectures. Nonvolatile memory devices, such as resistive random-access memory, ferroelectric field-effect transistor, and volatile static random-access memory devices, can be selected as synaptic devices. A multilayer perceptron and convolutional neural networks (CNNs), such as LeNet-5, VGG-16, and a custom CNN named C4W-1, are simulated to evaluate the effects of these synaptic devices on the training and inference outcomes. The dataset used in the simulations are MNIST, CIFAR-10, and a white blood cell dataset. By applying batch normalization and appropriate optimizers in the training phase, neuromorphic systems with very low-bit-width or binary weights could achieve high pattern recognition rates that approach software-based CNN accuracy. We also introduce spiking neural networks with RRAM-based synaptic devices for the recognition of MNIST handwritten digits.

We discuss first results from hidden photon dark matter searches made with a prototype of the Broadband Radiometric Axion/ALPs Search Setup (BRASS-p) in the range of particle mass of 49.63-74.44 $\mu$eV (frequency range of 12-18 GHz). The conceptual design of BRASS and a detailed description of its present prototype, BRASS-p, are given, with a view of the potential application of such setups to hidden photon, axion, and axion-like particle (ALP) dark matter searches using heterodyne detectors in the range of particle mass from 40$\mu$eV to 4000$\mu$eV (10 GHz to 1 THz). Pioneering measurements made with BRASS-p achieve the record sensitivity of (0.3--1.0)$\times$$10^{-13}$ to the kinetic mixing between the normal and hidden photons, assuming the dark matter is made entirely of unpolarized hidden photons. Based on these results, a discussion of further prospects for dark matter searches using the BRASS-p apparatus is presented., Comment: 17 pages, 13 figures. Prepared for submission to JCAP

Cameras and image-editing software often process images in the wide-gamut ProPhoto color space, encompassing 90% of all visible colors. However, when images are encoded for sharing, this color-rich representation is transformed and clipped to fit within the small-gamut standard RGB (sRGB) color space, representing only 30% of visible colors. Recovering the lost color information is challenging due to the clipping procedure. Inspired by neural implicit representations for 2D images, we propose a method that optimizes a lightweight multi-layer-perceptron (MLP) model during the gamut reduction step to predict the clipped values. GamutMLP takes approximately 2 seconds to optimize and requires only 23 KB of storage. The small memory footprint allows our GamutMLP model to be saved as metadata in the sRGB image -- the model can be extracted when needed to restore wide-gamut color values. We demonstrate the effectiveness of our approach for color recovery and compare it with alternative strategies, including pre-trained DNN-based gamut expansion networks and other implicit neural representation methods. As part of this effort, we introduce a new color gamut dataset of 2200 wide-gamut/small-gamut images for training and testing. Our code and dataset can be found on the project website: https://gamut-mlp.github.io.

Neural-based image and video codecs are significantly more power-efficient when weights and activations are quantized to low-precision integers. While there are general-purpose techniques for reducing quantization effects, large losses can occur when specific entropy coding properties are not considered. This work analyzes how entropy coding is affected by parameter quantizations, and provides a method to minimize losses. It is shown that, by using a certain type of coding parameters to be learned, uniform quantization becomes practically optimal, also simplifying the minimization of code memory requirements. The mathematical properties of the new representation are presented, and its effectiveness is demonstrated by coding experiments, showing that good results can be obtained with precision as low as 4~bits per network output, and practically no loss with 8~bits., Comment: 2022 IEEE International Conference on Image Processing (ICIP)

Partitioning algorithms play a key role in many scientific and engineering disciplines. A partitioning algorithm divides a set into a number of disjoint subsets or partitions. Often, the quality of the resulted partitions is measured by the amount of impurity in each partition, the smaller impurity the higher quality of the partitions. In general, for a given impurity measure specified by a function of the partitions, finding the minimum impurity partitions is an NP-hard problem. Let $M$ be the number of $N$-dimensional elements in a set and $K$ be the number of desired partitions, then an exhaustive search over all the possible partitions to find a minimum partition has the complexity of $O(K^M)$ which quickly becomes impractical for many applications with modest values of $K$ and $M$. Thus, many approximate algorithms with polynomial time complexity have been proposed, but few provide bounded guarantee. In this paper, an upper bound and a lower bound for a class of impurity functions are constructed. Based on these bounds, we propose a low-complexity partitioning algorithm with bounded guarantee based on the maximum likelihood principle. The theoretical analyses on the bounded guarantee of the algorithms are given for two well-known impurity functions Gini index and entropy. When $K \geq N$, the proposed algorithm achieves state-of-the-art results in terms of lowest approximations and polynomial time complexity $O(NM)$. In addition, a heuristic greedy-merge algorithm having the time complexity of $O((N-K)N^2+NM)$ is proposed for $K