Augmented Memory Research Articles

Optimization of In Vitro Th17 Polarization for Adoptive Cell Therapy in Chronic Lymphocytic Leukemia.

Although preclinical investigations have shown notable efficacy in solid tumor models utilizing in vitro-differentiated Th17 cells for adoptive cell therapy (ACT), the potential benefits of this strategy in enhancing ACT efficacy in hematological malignancies, such as chronic lymphocytic leukemia (CLL), remain unexplored. CLL is a B-cell malignancy with a clinical challenge of increased resistance to targeted therapies. T-cell therapies, including chimeric antigen receptor (CAR) T cells, have demonstrated limited success in CLL, which is attributed to CLL-mediated T-cell dysfunction and skewing toward immunosuppressive phenotypes. Herein, we illustrate the feasibility of polarizing CD4+ T cells from the Eμ-TCL1 murine model, the most representative model for human CLL, into Th17 phenotype, employing a protocol of T-cell activation through the inducible co-stimulator (ICOS) alongside a polarizing cytokine mixture. We demonstrate augmented memory properties of in vitro-polarized IL-17-producing T cells, and preliminary in vivo persistence in leukemia-bearing mice. Our findings gain translational relevance through successful viral transduction of Eμ-TCL1 CD4+ T cells with a CD19-targeted CAR construct during in vitro Th17 polarization. Th17 CAR T cells exhibited remarkable persistence upon encountering antigen-expressing target cells. This study represents the first demonstration of the potential of in vitro-differentiated Th17 cells to enhance ACT efficacy in CLL.

Simultaneous Multi-View Object Recognition and Grasping in Open-Ended Domains

To aid humans in everyday tasks, robots need to know which objects exist in the scene, where they are, and how to grasp and manipulate them in different situations. Therefore, object recognition and grasping are two key functionalities for autonomous robots. Most state-of-the-art approaches treat object recognition and grasping as two separate problems, even though both use visual input. Furthermore, the knowledge of the robot is fixed after the training phase. In such cases, if the robot encounters new object categories, it must be retrained to incorporate new information without catastrophic forgetting. To resolve this problem, we propose a deep learning architecture with an augmented memory capacity to handle open-ended object recognition and grasping simultaneously. In particular, our approach takes multi-views of an object as input and jointly estimates pixel-wise grasp configuration as well as a deep scale- and rotation-invariant representation as output. The obtained representation is then used for open-ended object recognition through a meta-active learning technique. We demonstrate the ability of our approach to grasp never-seen-before objects and to rapidly learn new object categories using very few examples on-site in both simulation and real-world settings. Our approach empowers a robot to acquire knowledge about new object categories using, on average, less than five instances per category and achieve 95%\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$95\\%$$\\end{document} object recognition accuracy and above 91%\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$91\\%$$\\end{document} grasp success rate on (highly) cluttered scenarios in both simulation and real-robot experiments. A video of these experiments is available online at: https://youtu.be/n9SMpuEkOgk

Construction and application of logistics scheduling model based on heterogeneous graph neural network

The core of logistics is scheduling and monitoring. After the modern interprise logistics development concept change, the development prospect of enterprise logistics is more optimistic. Major enterprises have begun to use intelligent logistics scheduling platforms. In order to solve the problem that heterogeneous information fusion is complex in the temporal heterogeneous graphs, this paper proposes to dynamically store and update node representation through an augmented memory matrix in a memory network. At the same time, the model also designs a novel read-write module for the memory matrix, which can effectively capture the timing information in the long interaction sequence and has high flexibility. The model has significantly improved in tasks such as node classification, timing recommendation and visualization. This paper studies the logistics supply chain of modern enterprises and establishes the mathematical model of vehicle scheduling. This paper takes the non-full load scheduling model as the critical research object. Based on the research of logistics supply chain, the vehicle scheduling model is established. The intelligent heuristic algorithm is applied to solve it, and the effective vehicle distribution scheme and driving route are formed. The simulation results show that the approximate Pareto optimal solution obtained by our designed model and algorithm has good robustness. NSGAIIROELSDR can get a better solution in small-scale scheduling. However, in large-scale numerical experiments, the final solution obtained by MOEA/DROELSDR is obviously better than that of NSGAIIROELSDR, and the running time of MOEA/DROELSDR is also shorter. Therefore, we conclude that MOEA/DROELSDR is more suitable for large-scale scheduling, and NSGAIIROELSDR is more suitable for more minor scheduling.

Recurrent unit augmented memory network for video summarisation

AbstractVideo summarisation can relieve the pressure on video storage, transmission, archiving, and retrieval caused by the explosive growth of online videos in recent years. Most existing supervised video summarisation methods use convolutional neural network (CNN) or recurrent neural network (RNN) to model the temporal dependencies between video frames or video shots. CNN mainly focuses on local information, and RNN loses long‐term information when the input sequence is long, both of which have limited ability to obtain long‐range memory in the video. Therefore, a recurrent unit augmented memory network (RUAMN) for video summarisation is proposed, which effectively utilises the long‐term memory extraction ability of the end‐to‐end memory network (MemN2N) and solves the problem that MemN2N is insensitive to temporal sequence information. At the same time, the proposed RUAMN enhances the process of memory update between multiple computational steps (hops), and finally generates a meaningful video summarisation result. Specifically, the proposed RUAMN is mainly composed of the input module, the global‐and‐local sampling, the memory module and the output module. The input module uses a bidirectional GRU to obtain the forward and backward information of each video frame. Then the global‐and‐local sampling performs global sampling and local sampling on the output sequence of the input module respectively to obtain several shorter sequences, so that the memory modules can capture the fine‐grained relationship features between video frames more effectively. The memory module extracts the long‐term memory information in the feature sequence, and finally the frame‐level importance scores are predicted by the output module. Extensive experiments on benchmark datasets, that is, TVSum and SumMe, demonstrate the superiority of our method over several state‐of‐the‐art supervised video summarisation methods.

The art of augmented memory: Remembering with augmented reality application in the case of the Temple of Artemis

Abstract This article focuses on the effects of different representation modes of architectural heritage in augmented reality (AR) applications on remembering. Deterioration of the tangible evidence of architectural heritage compromises not only its visibility in the heritage site, but also its presence in memory. Converging survived features and digitally produced representations of the heritage, AR applications in mobile devices provide the memory of the site with endurance, however what is remembered immensely depends on how the heritage is digitally represented on screen. Conceived as a case study for the method of analysis derived from classical memorizing technique of the art of memory, the ‘[AR]temis’ project, reported in this article, aimed to get insights into the effects of the representational qualities of augmented architectural heritage on remembering and also into future AR projects developed for architectural heritage sites with its original method of analysis to inform design decisions. The research project involved the development of the method of art of augmented memory, the AR application, as well as questionnaires and interviews with the respondents’ on-site tests of the application. The results of this analysis show that the decisions regarding the digital representation of architectural heritage in AR applications entail not only the visual qualities of the heritage per se, but also how the actual site of memory is visualized on screen.

Data Assimilation Network for Generalizable Person Re-Identification

In this paper, a data assimilation network is proposed to tackle the challenges of domain generalization for person re-identification (ReID). Most of the existing research efforts only focus on single-dataset issues, and the trained models are difficult to generalize to unseen scenarios. This paper presents a distinctive idea to improve the generality of the model by assimilating three types of images: style-variant images, misaligned images and unlabeled images. The latter two are often ignored in the previous domain generalization ReID studies. In this paper, a non-local convolutional block attention module is designed for assimilating the misaligned images, and an attention adversary network is introduced to correct it. A progressive augmented memory is designed for assimilating the unlabeled images by progressive learning. Moreover, we propose an attention adversary difference loss for attention correction, and a labeling-guide discriminative embedding loss for progressive learning. Rather than designing a specific feature extractor that is robust to style shift as in most previous domain generalization work, we propose a data assimilation meta-learning procedure to train the proposed network, so that it learns to assimilate style-variant images. It is worth mentioning that we add an unlabeled augmented dataset to the source domain to tackle the domain generalization ReID tasks. Extensive experiments demonstrate that our approach significantly outperforms the state-of-the-art domain generalization methods.

Learning Augmented Memory Joint Aberrance Repressed Correlation Filters for Visual Tracking

With its outstanding performance and tracking speed, discriminative correlation filters (DCF) have gained much attention in visual object tracking, where time-consuming correlation operations can be efficiently computed utilizing the discrete Fourier transform (DFT) with symmetric properties. Nevertheless, the inherent issues of boundary effects and filter degradation, as well as occlusion and background clutter, degrade the tracking performance. In this work, we proposed an augmented memory joint aberrance repressed correlation filter (AMRCF) for visual tracking. Based on the background-aware correlation filter (BACF), we introduced adaptive spatial regularity to mitigate the boundary effect. Several historical views and the current view are exploited to train the model together as a way to reinforce the memory. Furthermore, aberrance repression regularization was introduced to suppress response anomalies due to occlusion and deformation, while adopting the dynamic updating strategy to reduce the impact of anomalies on the appearance model. Finally, extensive experimental results over four well-known tracking benchmarks indicate that the proposed AMRCF tracker achieved comparable tracking performance to most state-of-the-art (SOTA) trackers.

Memory augmented recurrent neural networks for de-novo drug design.

A recurrent neural network (RNN) is a machine learning model that learns the relationship between elements of an input series, in addition to inferring a relationship between the data input to the model and target output. Memory augmentation allows the RNN to learn the interrelationships between elements of the input over a protracted length of the input series. Inspired by the success of stack augmented RNN (StackRNN) to generate strings for various applications, we present two memory augmented RNN-based architectures: the Neural Turing Machine (NTM) and the Differentiable Neural Computer (DNC) for the de-novo generation of small molecules. We trained a character-level convolutional neural network (CNN) to predict the properties of a generated string and compute a reward or loss in a deep reinforcement learning setup to bias the Generator to produce molecules with the desired property. Further, we compare the performance of these architectures to gain insight to their relative merits in terms of the validity and novelty of the generated molecules and the degree of property bias towards the computational generation of de-novo drugs. We also compare the performance of these architectures with simpler recurrent neural networks (Vanilla RNN, LSTM, and GRU) without an external memory component to explore the impact of augmented memory in the task of de-novo generation of small molecules.

Latent anxiety in clinical depression is associated with worse recognition of emotional stimuli

Background: Major Depressive Disorder, characterized by cognitive affective biases, is a considerable public health challenge. Past work has shown that higher depressive symptoms are associated with augmented memory of negative stimuli. In contrast, anxiety symptoms have been associated with overgeneralization of emotional memories. Given the high comorbidity of depression and anxiety, it is critical to understand how cognitive affective biases are differentially associated with clinical symptoms.Method: We used continuous measures of depression (Beck Depression Inventory [BDI-II]) and anxiety (Beck Anxiety Inventory [BAI]) to evaluate an adult sample (N = 79; 18–41 years old, 58 female). Emotional memory discrimination and recognition memory were tested using an emotional discrimination task. We applied exploratory factor analysis to questions from the BAI and BDI-II to uncover latent constructs consisting of negative affect, anhedonia, somatic anxiety, and cognitive anxiety.Results: We report evidence that anxious symptoms were associated with impaired recognition of negative items after accounting for age and sex. Our exploratory factor analysis revealed that impaired negative item recognition is largely associated with somatic and cognitive anxiety factors.Limitations: Interpretations in a mixed pathology sample, especially given collinearity among factors, may be difficult.Conclusions: We provide evidence that somatic and cognitive anxiety are related to impaired recognition memory for negative stimuli. Future clinical investigations should uncover the neurobiological basis supporting the link between recognition of negative stimuli and somatic/cognitive symptoms of anxiety.

On the memory benefits of repeated study with variable tasks.

Encoding variability refers to the situation in which repeated items are processed in different ways on each presentation. Superior memory performance resulting from encoding variability is sometimes argued to underlie important phenomena in human memory such as the spacing effect. However, the memory benefits of encoding variability are often elusive. Here we investigated encoding variability in ten experiments in which participants studied words with the same or different orienting tasks across repetitions. We have found the benefits of variable encoding to depend on the number of learning cycles and the retrieval demands at test. These results are interpreted in light of a distinction between different components of memory representations established at study, suggesting that encoding variability promoted via different orienting tasks-as implemented in the present study-fosters more elaborate encoding of semantic features. This augmented semantic component benefits memory performance only when a memory test is used that taps predominantly semantic features of memory representations, minimizing the role of contextual and relational factors. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

LVD-NMPC: A learning-based vision dynamics approach to nonlinear model predictive control for autonomous vehicles

In this article, we introduce a learning-based vision dynamics approach to nonlinear model predictive control (NMPC) for autonomous vehicles, coined learning-based vision dynamics (LVD) NMPC. LVD-NMPC uses an a-priori process model and a learned vision dynamics model used to calculate the dynamics of the driving scene, the controlled system’s desired state trajectory, and the weighting gains of the quadratic cost function optimized by a constrained predictive controller. The vision system is defined as a deep neural network designed to estimate the dynamics of the image scene. The input is based on historic sequences of sensory observations and vehicle states, integrated by an augmented memory component. Deep Q-learning is used to train the deep network, which once trained can also be used to calculate the desired trajectory of the vehicle. We evaluate LVD-NMPC against a baseline dynamic window approach (DWA) path planning executed using standard NMPC and against the PilotNet neural network. Performance is measured in our simulation environment GridSim, on a real-world 1:8 scaled model car as well as on a real size autonomous test vehicle and the nuScenes computer vision dataset.

Analysis of lifting scheme based Double Density Dual-Tree Complex Wavelet Transform for de-noising medical images

Medical images play a vital role in diagnosis of various diseases. This has paved a path to the extensive use of CT, mammogram, MRI and ultrasound images in the recent days which has caused a rising concern about the radiation dosage that is involved in medical screening process. Owing to this concern low dose screening is widely being performed and has resulted in the introduction of noise, artifacts thus producing low image quality which can adversely affect the judgment of the radiologists. This in turn has led to the demand of enhanced image de-noising techniques. This work is an approach to remove multiple types of noises from low dose medical images using lifting based Double Density Dual-Tree Complex Wavelet Transform (DDDTCWT) and a modified Bernoulli based thresholding technique enhanced by fuzzy optimization technique. The parameters observed from the simulation results of the proposed method were compared with the existing de-noising techniques and results of the proposed method have shown significant improvement over the conventional techniques. The proposed work not only efficiently de-noises the image but also enhances its visual appearance. The Lifting scheme used provides augmented memory for decomposition, thus speeding up the entire de-noising process.

Loss of the orphan nuclear receptor NR2F6 enhances CD8+ T-cell memory via IFN-\u03b3

Memory formation is a hallmark of T cell-mediated immunity, but how differentiation into either short-lived effector cells (SLECs, CD127−KLRG1+) or memory precursors cells (MPECs, CD127+KLRG1−) and subsequent regulation of long-term memory is adjusted is incompletely understood. Here, we show that loss of the nuclear orphan receptor NR2F6 in germ-line Nr2f6-deficient mice enhances antigen-specific CD8+ memory formation up to 70 days after bacterial infection with Listeria monocytogenes (LmOVA) and boosts inflammatory IFN-γ, TNFα, and IL-2 cytokine recall responses. Adoptive transfer experiments using Nr2f6−/− OT-I T-cells showed that the augmented memory formation is CD8+ T-cell intrinsic. Although the relative difference between the Nr2f6+/+ and Nr2f6−/− OT-I memory compartment declines over time, Nr2f6-deficient OT-I memory T cells mount significantly enhanced IFN-γ responses upon reinfection with increased clonal expansion and improved host antigen-specific CD8+ T-cell responses. Following a secondary adoptive transfer into naïve congenic mice, Nr2f6-deficient OT-I memory T cells are superior in clearing LmOVA infection. Finally, we show that the commitment to enhanced memory within Nr2f6-deficient OT-I T cells is established in the early phases of the antibacterial immune response and is IFN-γ mediated. IFN-γ blocking normalized MPEC formation of Nr2f6-deficient OT-I T cells. Thus, deletion or pharmacological inhibition of NR2F6 in antigen-specific CD8+ T cells may have therapeutic potential for enhancing early IFN-γ production and consequently the functionality of memory CD8+ T cells in vivo.

Augmented Memory Replay in Reinforcement Learning With Continuous Control

Online reinforcement learning agents are currently able to process an increasing amount of data by converting it into a higher order value functions. This expansion of the information collected from the environment increases the agent's state space enabling it to scale up to a more complex problems but also increases the risk of forgetting by learning on redundant or conflicting data. To improve the approximation of a large amount of data, a random mini-batch of the past experiences that are stored in the replay memory buffer is often replayed at each learning step. The proposed work takes inspiration from a biological mechanism which act as a protective layer of human brain higher cognitive functions: active memory consolidation mitigates the effect of forgetting of previous memories by dynamically processing the new ones. The similar dynamics are implemented by a proposed augmented memory replay AMR capable of optimizing the replay of the experiences from the agent's memory structure by altering or augmenting their relevance. Experimental results show that an evolved AMR augmentation function capable of increasing the significance of the specific memories is able to further increase the stability and convergence speed of the learning algorithms dealing with the complexity of continuous action domains.

Binarization Methods for Motor-Imagery Brain–Computer Interface Classification

Successful motor-imagery brain-computer interface (MI-BCI) algorithms either extract a large number of handcrafted features and train a classifier, or combine feature extraction and classification within deep convolutional neural networks (CNNs). Both approaches typically result in a set of real-valued weights, that pose challenges when targeting real-time execution on tightly resource-constrained devices. We propose methods for each of these approaches that allow transforming real-valued weights to binary numbers for efficient inference. Our first method, based on sparse bipolar random projection, projects a large number of real-valued Riemannian covariance features to a binary space, where a linear SVM classifier can be learned with binary weights too. By tuning the dimension of the binary embedding, we achieve almost the same accuracy in 4-class MI (≤1.27% lower) compared to models with float16 weights, yet delivering a more compact model with simpler operations to execute. Second, we propose to use memory-augmented neural networks (MANNs) for MI-BCI such that the augmented memory is binarized. Our method replaces the fully connected layer of CNNs with a binary augmented memory using bipolar random projection, or learned projection. Our experimental results on EEGNet, an already compact CNN for MI-BCI, show that it can be compressed by $1.28\times $ at iso-accuracy using the random projection. On the other hand, using the learned projection provides 3.89% higher accuracy but increases the memory size by $28.10\times $ .

Alarmin-painted exosomes elicit persistent antitumor immunity in large established tumors in mice

Treating large established tumors is challenging for dendritic cell (DC)-based immunotherapy. DC activation with tumor cell-derived exosomes (TEXs) carrying multiple tumor-associated antigen can enhance tumor recognition. Adding a potent adjuvant, high mobility group nucleosome-binding protein 1 (HMGN1), boosts DCs’ ability to activate T cells and improves vaccine efficiency. Here, we demonstrate that TEXs painted with the functional domain of HMGN1 (TEX-N1ND) via an exosomal anchor peptide potentiates DC immunogenicity. TEX-N1ND pulsed DCs (DCTEX-N1ND) elicit long-lasting antitumor immunity and tumor suppression in different syngeneic mouse models with large tumor burdens, most notably large, poorly immunogenic orthotopic hepatocellular carcinoma (HCC). DCTEX-N1ND show increased homing to lymphoid tissues and contribute to augmented memory T cells. Importantly, N1ND-painted serum exosomes from cancer patients also promote DC activation. Our study demonstrates the potency of TEX-N1ND to strengthen DC immunogenicity and to suppress large established tumors, and thus provides an avenue to improve DC-based immunotherapy.

Memory Reorganization: A Symmetric Memory Network for Reorganizing Neighbors and Topics to Complete Rating Prediction

Using pre-trained topic information to assist in training neural networks can effectively support the completion of the rating prediction task. However, existing neural-topic methods consider only the use of topic information corresponding to current users and items without neighbors, whereas existing memory-based neighborhood approaches are inappropriate for the direct modeling of neighbors with topics. To address the limitations, we argue that memory networks have the ability to organize neighbors with corresponding topics well and can provide a general solution to this problem. To confirm our hypothesis, we propose two approaches. One is an augmented memory network to couple with and enhance existing neural-topic models. The other is a symmetric memory network activated by a memory reorganization mechanism, which is a compact and generalized method for rating prediction. The experimental results demonstrate the effectiveness of the memory reorganization mechanism and show that the two proposed methods have advantages over existing state-of-the-art topic modeling approaches.

“The Project Changed My Life”: Sport Education’s Transformative Potential on Student Physical Literacy

ABSTRACTPurpose: This longitudinal study provided a retrospective examination of the impact of student participation in a year-long Sport Education curricular experience on physical literacy development. A focus was placed on students’ reported motivation, attitudes and disposition to participate in physical education (PE) and sport, and their re-actualizations of a healthy sport culture in the years following the Sport Education experience. Methods: Three years following the Sport Education experience, a cohort of 24 tenth-grade students (nine girls and 15 boys) were traced back and invited to share their recollected memories and accounts of their PE experience and subsequent attributes of physical literacy. The data collection protocol included augmented memory retrieval techniques and a retrospective survey, which informed semi-structured individual and focus groups interviews. A thematic analysis of data was performed and trustworthiness criteria included data triangulation, crosschecking, and collaborative interpretational analysis. Results: Prolonged participation in a year-long Sport Education curriculum promoted a transformation in students’ physical literacy dispositions, both within and outside the school context. Physical literacy attributes included empathetic attitudes toward others, resilience against discriminatory learning activities and a commitment of the participants to actively pursue equitable and meaningful participation in PE and sport. Embodied interpersonal, managerial, and peer-teaching skills sustained the participants’ sport culture activism against impoverished instructional practices in PE. Conclusions: Prolonged participation in Sport Education, and consequential development of enthusiasm, competency, and literacy can positively contribute to individuals’ lifelong physical literacy journey.

AI learns how to learn with TCAMs

Neural networks could learn new concepts quickly and from only a few examples by using a ferroelectric ternary content-addressable memory as an augmented memory.

GAMENet: Graph Augmented MEmory Networks for Recommending Medication Combination

Recent progress in deep learning is revolutionizing the healthcare domain including providing solutions to medication recommendations, especially recommending medication combination for patients with complex health conditions. Existing approaches either do not customize based on patient health history, or ignore existing knowledge on drug-drug interactions (DDI) that might lead to adverse outcomes. To fill this gap, we propose the Graph Augmented Memory Networks (GAMENet), which integrates the drug-drug interactions knowledge graph by a memory module implemented as a graph convolutional networks, and models longitudinal patient records as the query. It is trained end-to-end to provide safe and personalized recommendation of medication combination. We demonstrate the effectiveness and safety of GAMENet by comparing with several state-of-the-art methods on real EHR data. GAMENet outperformed all baselines in all effectiveness measures, and also achieved 3.60% DDI rate reduction from existing EHR data.