Recognition Function Research Articles

Gesture-Driven Home Automation System

This project addresses the critical challenge of making home automation accessible and inclusive for individuals with physical disabilities and visual impairments. Existing home automation methods, such as voice control, mobile apps, and timers, exhibit shortcomings that hinder their effectiveness for this demographic. To overcome these limitations, the project shows the development of a gesture-based home automation system. This system integrates gesture recognition and security functionality. It operates in two stages: user authentication through face recognition and gesture-based appliance control. A Convolutional Neural Network (CNN) processes webcam-captured gestures, with signals relayed to an Arduino via Pyfirmata for appliance activation. A model was trained where it could detect the human hand(s) and more soshow the 20 landmarks that media pipe offers. The system worked well in controlling various home appliances using gestures. Through this innovative approach, the project contributes to the advancement of accessible smart home technology, fostering independence and inclusivity for users with diverse abilities.

The Influence of Vowels on the Identification of Spoken Disyllabic Words in the Malayalam Language for Individuals with Hearing Loss

Background/Objectives: The present study investigates the reasons for better recognition of disyllabic words in Malayalam among individuals with hearing loss. This research was conducted in three experiments. Experiment 1 measured the psychometric properties (slope, intercept, and maximum scores) of disyllabic wordlists. Experiment 2 examined PBmax scores across varying degrees of sensorineural hearing loss (SNHL) and compared these findings with studies in other Indian and global languages. Experiment 3 analyzed the recognition performance of different vowel combinations across varying degrees of hearing loss. Methods: Experiment 1: Psychometric functions for disyllabic word recognition were derived from 45 individuals with normal hearing. Word recognition was tested in quiet at nine hearing levels ranging from −10 to +40 dB HL. Experiment 2: 1000 participants with SNHL were categorized by hearing loss severity (mild, moderate, moderately severe, severe, and profound). Word recognition scores, including PBmax, were analyzed and compared across severity levels. Experiment 3: Percent error scores for 17 vowel combinations were assessed in 37 participants with SNHL. Ten disyllabic words represented each combination. Results: Disyllabic wordlists showed significantly higher word recognition scores than monosyllabic lists across all degrees of hearing loss. Individuals with mild-to-moderately severe SNHL achieved higher PBmax scores, with performance declining at severe- and profound-loss levels. The higher recognition of disyllabic words was attributed to contextual cues and low-frequency vowel-based information, particularly benefiting those with residual low-frequency hearing. Error analysis highlighted the influence of specific vowel combinations on word recognition performance. Conclusions: Disyllabic words are easier to recognize than monosyllabic words for individuals with SNHL due to their rich contextual and low-frequency energy cues. Disyllabic wordlists sustain higher recognition scores up to moderately severe hearing loss but show a marked decline with more severe losses. The phonemic balance of wordlists and vowel combinations significantly influences word recognition, emphasizing the importance of these factors in developing wordlists for clinical use.

Abnormal tRNA epigenetic modifications and related impact on neurodegenerative diseases.

Neurodegenerative diseases are a heterogeneous group of neurological disorders characterized by the progressive loss of neurons in the central or peripheral nervous system. Research on the pathogenesis and drug targets of these diseases still faces many challenges due to the complex etiology. In recent years, the role of transfer RNA (tRNA) epigenetic modifications in neurodegenerative diseases has attracted widespread attention. The tRNA modification is crucial for regulating codon recognition, maintaining molecular structural stability, and the generation of tRNA-derived fragments (tRFs). Recent studies have highlighted a close association between abnormal tRNA modifications and the pathogenesis of various neurodegenerative diseases; especially for abnormalities of elongator complex-dependent tRNA modification and methylation modification, which impact the translation process and tRFs levels. These changes regulate protein homeostasis and cellular stress responses, ultimately influencing the survival of neuronal cells. Moreover, significant changes in tRFs levels have been noted in neurodegenerative diseases, and special tRFs show distinct effects on neurodegenerative diseases. This review aims to provide an overview of the physiological functions of tRNA epigenetic modifications and their regulatory mechanisms in neurodegenerative diseases, covering both classical functions such as codon recognition and non-classical functions such as tRFs biogenesis. Additionally, the potential of targeting tRNA modifications for therapeutic applications is also discussed.

Hypoimmunogenic CD19 CAR-NK cells derived from embryonic stem cells suppress the progression of human B-cell malignancies in xenograft animals

BackgroundChimeric antigen receptor (CAR) engineered natural killer (NK) cells exhibit advantages such as MHC-independent recognition and strong anti-tumor functions. However, allogeneic CAR-NK cells derived from human tissues are heterogeneous and susceptible to clearance by hosts.MethodsWe generated a B2M knockout, HLA-E and CD19 CAR ectopic expressing embryonic stem cell (ESC) line, which differentiated normally and gave rise to homogeneous CD19 CAR-NK (CD19 CAR-UiNK) cells using an organoid aggregate induction method. The CD19 CAR-UiNK were co-cultured with T cells or NK cells derived from peripheral blood mononuclear cells (PBMC) with the mismatched HLA to evaluate the immunogenicity of CD19 CAR-UiNK cells. We further assessed the therapeutic effects of CD19 CAR-UiNK cells on CD19+ tumor cells through in vitro cytotoxicity assays and in vivo animal models.ResultsThe CD19 CAR-UiNK cells exhibited typical expression patterns of activating and inhibitory receptors, and crucial effector molecules of NK cells, similar to those of unmodified NK cells. In co-culture assays, the CD19 CAR-UiNK cells evaded allogeneic T cell response and suppressed allogeneic NK cell response. Functionally, the CD19 CAR-UiNK cells robustly secreted IFN-γ and TNF-α, and upregulated CD107a upon stimulation with Nalm-6 tumor cells. The CD19 CAR-UiNK cells effectively eliminated CD19+ tumor cells in vitro, including B-cell cancer cell lines and primary tumor cells from human B-cell leukemia and lymphoma. Further, the CD19 CAR-UiNK cells exhibited strong anti-tumor activity in xenograft animals.ConclusionWe offer a strategy for deriving homogeneous and hypoimmunogenic CD19 CAR-iNK cells with robust anti-tumor effects from ESCs. Our study has significant implications for developing hypoimmunogenic CD19 CAR-NK cell therapy using human ESC as an unlimited cell source.

DNA Nanolock-Based Logic Gate-Directed Reciprocal Feedback for Stepwise Cell Typing and Combination Treatment.

To achieve accurate molecular diagnosis and early-stage intervention of disease on demand, there is an urgent need for the monitoring of multiple biomarkers and multipath information acquisition in living cells. The DNA combinatorial logic gate is an appropriate strategy for providing a systematic proof of concept with comprehensive information and function. Herein, a modular DNA logic gate nanomachine is designed for sufficient multistep reciprocal cell identification and therapy via the iteration of simple logic operations. In this logic gate system, this main module is constructed by G-quadruplex-locked gold nanocages (AuNCs), serving dual functions of drug encapsulation and cell recognition. The logic system is composed of OR, XNOR, AND, and NOR gates employing two intracellular disease biomarkers (microRNA 21 and microRNA 155) as inputs and the fluorescence signal of doxorubicin (Dox) as an output. The output signals of the four logic gates are iterated to process the imaging analysis data from the complex matrix in the living cell. Via positive and negative reciprocal feedback, the series circuit of different gates enables different functions, including the preliminary screening and the distinction of the cell type. Through the mutual preliminary screening and further proof, this logic system achieves accurate identification of cells, controlled drug release, and photothermal treatment using the AuNC as a photothermal transducer. This DNA logic system broadens the applications of the biocomputing system in disease screening and logic-controlled treatment fields.

When France Advised Syria: Cooperation as a Principle of Recognition, and its Limits

Abstract This article is a re-examination of cooperation in international relations through the example of Franco-Syrian relations between 2001 and 2004. I explore how cooperation fulfils – or not – a function of recognition, especially in the case of countries under unequal terms. Based on archival research and interviews, the article analyses three concrete moments of cooperation between Syria and France during this time. While traditional Franco-Syrian cooperation, prior to 2000, was based on a mutual recognition of partners, in the aftermath of presidential change in Syria in 2000 and within the shifting regional context, the Franco-Syrian relationship seemed to be marked by an inequality. Therefore, the article aims to capture the symbols of this recognition between France and Syria, and to study the role of cooperation not only in deepening a bilateral relationship, but – in a new perspective in international relations – in undermining it.

Evaluating a generative artificial intelligence accuracy in providing medication instructions from smartphone images

BackgroundThe Food and Drug Administration mandates patient labeling materials like the Medication Guide (MG) and Instructions for Use (IFU) to support appropriate medication use. However, challenges such as low health literacy and difficulties navigating these materials may lead to incorrect medication usage, resulting in therapy failure or adverse outcomes. The rise of generative AI, presents an opportunity to provide scalable, personalized patient education through image recognition and text generation. ObjectiveThis study aimed to evaluate the accuracy and safety of medication instructions generated by ChatGPT based on user-provided drug images, compared to the manufacturer's standard instructions. MethodsImages of 12 medications requiring multiple steps for administration were uploaded to ChatGPT's image recognition function. ChatGPT's responses were compared to the official IFU and MG using text classifiers, Count Vectorization (CountVec), and Term Frequency-Inverse Document Frequency (TF-IDF). The clinical accuracy was further evaluated by independent pharmacists to determine if ChatGPT responses were valid for patient instruction. ResultsChatGPT correctly identified all medications and generated patient instructions. CountVec outperformed TF-IDF in text similarity analysis, with an average similarity score of 76%. However, clinical evaluation revealed significant gaps in the instructions, particularly for complex administration routes, where ChatGPT's guidance lacked essential details, leading to lower clinical accuracy scores. ConclusionWhile ChatGPT shows promise in generating patient-friendly medication instructions, its effectiveness varies based on the complexity of the medication. The findings underscore the need for further refinement and clinical oversight to ensure the safety and accuracy of AI-generated medical guidance, particularly for medications with complex administration processes.

Prenatal ethanol exposure impairs hippocampal plasticity and cognition in adolescent mice

BackgroundPrenatal alcohol exposure (PAE) induces a wide range of neurodevelopmental disabilities that are grouped under the term ‘fetal alcohol spectrum disorders’ (FASD). The effects of PAE on brain development are dependent on complex neurochemical events, including modification of AMPA receptors (AMPARs). We have recently found that chronic ethanol (EtOH) exposure decreases AMPA-mediated neurotransmission and expression through the overexpression of the specific microRNA (miR)137 and 501-3p, which target GluA1 AMPA subunit, in the developing hippocampus in vitro. Here, we explored how PAE mice may alter AMPAergic synapses in the hippocampus, and its effects on behavior. MethodsTo model PAE, we exposed C57Bl/6 pregnant mice to 10 % EtOH during during the first 10 days of gestation (GD 0–10; equivalent to the first trimester of pregnancy in humans). AMPA subunits postsynaptic expression in the hippocampus, electrical properties of CA1 neurons, memory recognition, and locomotor functions were then analyzed in adolescent PAE-exposed offspring. ResultsPAE adolescent mice showed dysregulation of AMPAergic neurotransmission, and increased miR 501-3p expression, associated with a significant reduction of spontaneous AMPA currents and intrinsic somatic excitability. In addition, PAE reduced the phosphorylation of AMPAR-containing GluA1 subunit, despite an increase in its total levels. Of note, the total levels of GluA2 and GluA3 AMPA receptors were enhanced as well. Consistently, at behavioral level, PAE reduced object recognition without altering locomotor activity. ConclusionsOur study shows that PAE leads to dysfunctional formation of AMPAergic synapses that could be responsible for neurobehavioral impairments, contributing to the understanding of the pathogenesis of FASD.

Wrinkled Rhenium Disulfide for Anisotropic Nonvolatile Memory and Multiple Artificial Neuromorphic Synapses.

Two-dimensional materials are emerging as potential solutions for high-density nonvolatile memory and efficient neuromorphic computing. However, integrating multidimensional memory and an ideal linear weight updating synapse in a simple device configuration to achieve versatile biomimetic neuromorphic systems remains challenging. Here, we introduce a wrinkled rhenium disulfide (ReS2) transistor, where the wrinkled structure facilitates the carrier trapping/detrapping at the dielectric interface, thus enabling the fusion of nonvolatile memory and both electronic and optoelectronic synaptic functionalities. As a nonvolatile memory, anisotropic wrinkled ReS2 can yield three distinct sets of data across three crystal orientations under identical programming operations. Each set demonstrates exceptional retention and endurance properties. As a neuromorphic synapse, it realizes the linear and symmetric updates of conductance states up to 9 bits and 8 bits, the ultra-low-energy consumption of 75 fJ and 2.5 pJ under the electrical and optical stimuli, respectively. The artificial neural network (ANN) based on electronic synapses gives a superior recognition accuracy of 92.9% for the original handwritten digits. The anisotropic synaptic responses and multiwavelength sensitivities of optoelectronic synapses enable them to execute advanced memory and recognition functions for complex images that encompass a variety of pattern features or color information. This underscores its substantial potential for integration into efficient biomimetic visual systems.

A multilevel resistive switching memristor based on flexible organic–inorganic hybrid film with recognition function

Abstract Brain-inspired neuromorphic computing systems fueled the emergence of memristor-based artificial synapses, however, conventional silicon-based devices restricted their usage in the wearable field because of their difficulty in bending. To tackle the above challenge, a vertically structured flexible memristor with aluminum-based hydroquinone organic–inorganic hybrid film and Al2O3 as the functional layer, ITO and Pt as the bottom and top electrodes, and PET as the substrate has been developed utilizing molecular/atomic layer deposition to achieve a tradeoff between the resistive transition properties and the flexibility of memristors. The obtained devices combine stable resistive switching behavior and flexibility, showing high switching ratio of 103, better retention (up to 105 s) and endurance properties (up to 104 cycles), and robustness at radius of curvature of 4.5 mm after 104 bending cycles. Furthermore, the presence of multilevel resistive states in these devices ensures that the memristor can emulate synaptic properties such as paired-pulse facilitation, transition from short-term plasticity to long-term plasticity, long-term potentiation and depression, and spike-time-dependent plasticity. The resistive switching mechanism and the role of the bending state on the electrical performance of the device are explored. The fully connected artificial neural network based on the memristor can achieve a recognition accuracy of 90.2% for handwritten digits after training and learning. Flexible memristor will bring feasible advances to the integration of neuromorphic computing and wearable functionality.

Establishment and application of autoverification system for HbA1c testing.

This study aimed to determine autoverification rules for routine glycated hemoglobin (HbA1c) analysis based on high-performance liquid chromatography (HPLC) principle. Laboratory information system (LIS) and Bio-Rad D-100 Advisor software (Bio-Rad, Hercules, USA) with graphics recognition function were carriers for the autoverification system. A total of 105,126 HbA1c results, including 98,249 HbA1c matching fast plasma glucose (FPG) results of real-world data from May 2019 to June 2020, were collected to determine autoverification rules including flags, delta checks, reporting limits, and logical rules. The validation database was composed of 48,045 HbA1c results and 41,083 matching FPG results. Autoverification passing rate and the reduction of turnaround time (TAT) were evaluated. Four autoverification systems (A, B, C, D) were established by two types of delta check rules, 28 flags, one reporting limits, and two kinds of logical rules. The autoverification passing rates were 80.6%, 78.8%, 83.7%, and 81.3%, and the average time saved in TAT were 117.5 min, 116.7 min, 121.1 min, and 121.7 min, respectively. Autoverification system C was the optimal one. Application of distribution of FPG corresponding to HbA1c groups had better performance as logical rules. Established HbA1c autoverifcation system shortened the auditing report time and improved work efficiency.

A study of long-term supraphysiologic-dose anabolic-androgenic steroid use on cognitive function in middle-aged men

ABSTRACT Background: Long-term use of supraphysiologic doses of anabolic-androgenic steroids (AAS) has been associated with impaired visuospatial memory in young men but little is known about its cognitive effects in middle-aged men. Objectives: We compared cognition in middle-aged men with histories of long-term AAS use and age-matched non-users. Methods: We administered cognitive tests from the CANTAB battery to 76 weightlifters aged 37–60 years (mean [SD] 48.5 [6.5] years), of whom 51 reported at least 2 years of cumulative AAS use and 25 reported no AAS exposure. Results: We found no significant AAS user versus non-user group differences on visuospatial, verbal memory, emotional recognition, or executive function tasks (corrected p’s ≥ .00089; effect sizes ≤ .5). Conclusions: Our null visuospatial task findings contrast with our prior younger cohort study (mean age 37.1 [7.1] years), in which we found impaired visuospatial task performance in people who use AAS, and with other reports of cognitive impairments in younger men use AAS. Men who use AAS may develop early visuospatial memory deficits that stabilize by middle age while middle-aged non-users’ performance may “catch up” due to normal age-related visuospatial declines. Similar effects could contribute to our null findings on other tasks. Between-study cohort substance use differences or environmental factor differences that modify cognition, such as study geographical location and time of year, also could contribute to our discordant findings. Since young adult male AAS users experience increased mortality from unnatural causes, improving our understanding of AAS cognitive effects in this age group is important.

Digital health apps : How to achieve the successful integration into healthcare

Digital health applications (apps) have been available on prescription since 2019 and offer a multitude of functions in health monitoring, symptom recognition, follow-up monitoring and patient care. The Digital Care Act from 2023 strives for a comprehensive integration of apps into medical care. The utilization of artificial intelligence (AI) in health apps promises an earlier diagnosis, sensitive symptom monitoring, optimization of courses of treatment, more effective doctor-patient communication and saving of resources. The integration of health apps and AI can enable a stronger personalized high-value care.

Simultaneous imaging of telomerase activity and protein tyrosine kinase 7 in living cells during epithelial-mesenchymal transformation via a near-infrared light-activatable nanoprobe

Exploring the relationship between key regulation molecules (such as telomerase and protein tyrosine kinase 7) during epithelial-mesenchymal transformation of cells is beneficial for studying malignant tumor metastasis. Fluorescence is usually used for real-time monitoring the distribution and expression of regulatory molecules in living cells. However, the recognition function of these classical nanoprobes is “always active” due to the absence of exogenous control, which leads to the amplification of both the background signal and the response signal, making it difficult to distinguish changes in biomolecule expression levels. To improve the fluorescence ratio between tumor and normal cells, we constructed near-infrared light-activatable nanoprobes by engineering the functional units of catalytic hairpin assembly and integrating upconversion luminescence nanoparticles. Under near-infrared light irradiation, the nanoparticles, serving as a near-infrared-to-ultraviolet light transducer, induced the photolysis of the photo-cleavable linkers sealed in hairpins. The recognition function of the nanoprobes can be controlled by near-infrared light, preventing them from recognizing the targets in non-irradiated regions. By employing the nanoprobes, we realized simultaneous imaging of two regulatory molecules in living cells and observed an increase in telomerase activity and a decrease in protein tyrosine kinase 7 expression during drug-induced epithelial-mesenchymal transformation. This work provides a promising method for revealing changes and relationships of regulatory molecules during tumor metastasis.

Targeted recognition and enhanced biotransformation of phytochemicals by a dual-functional cellulose-based hydrogel bioreactor

The combined technology of molecular imprinting and immobilization is a promising strategy for improving biocatalytic performance. In this work, a dual-functional cellulose-based hydrogel bioreactor with targeted recognition and transformation functions was innovatively designed by cellulose-based molecularly imprinted polymers (CM-MIPs) hydrogel coupled with layered double hydroxide immobilized enzyme (LDH-E). Firstly, CM-MIPs hydrogel was prepared by using phlorizin, 4-pinylpyridine, and cellulose microspheres as template molecule, functional monomer, and support material, respectively. Meanwhile, layered double hydroxide (LDH) taken as a carrier to immobilize β-glucosidase. The prepared LDH was layered sheet-like structure with ultra-thin thickness of approximately only 1.5 nm, and β-glucosidase was immobilized on both sides of the LDH. Further, the dual-functional bioreactor was constructed by the anion exchange, which possessed maximum targeted adsorption capacity to phlorizin of 15.25 mg/g, exhibited 1.2 folds higher transformation efficiency than LDH-E, and the phloretin content increased 26 folds to that in Lithocarpus litseifolius (Hance) Chu leaves extracts. Moreover, the transformation efficiency remained above 70 % even after five consecutive transformations. Overall, the dual-functional bioreactor has broad prospects for the application in targeted recognition and transformation of phytochemicals, and provides a new insight on multifunctional bio-based reactors in natural production field.

Relation between Emotional Recognition, Attachment Styles and Social Functioning among Psychiatric Patients

Relation between Emotional Recognition, Attachment Styles and Social Functioning among Psychiatric Patients

Design and Implementation of Gesture Music Composition System based on RealSense and Recurrent Neural Network

Aiming at the problems of slow human-computer interaction and low accuracy of hand gesture recognition (HGR) in music composition systems such as human-computer interaction, gesture recognition, and algorithmic composition, this paper designs an HGR music composition system based on real-sense technology and recurrent neural network. The system uses an Intel RealSense 3D camera to shoot the user's hand and extract the information on the hand joint points with Smart Wearable Biosensors (SWBSs). The joint points are used to construct a three-dimensional model of the hand bone, and the joint points' equivalent distance and joint deflection are calculated by the joint point information to match the corresponding Curwen HGR. The matching music data is input into the recurrent neural network. GRU is combined with the Markov chain algorithm to complete the composition and avoid the phenomenon of gradient disappearance or gradient explosion. The HGR recognition function and composition algorithm are simulated and tested to verify the method's feasibility. Cohort verification investigations and efficacy evaluations of wearable biosensors are essential to support their clinical acceptability. The research results show that the HGR music composition system based on real-sense technology and recurrent neural networks can effectively identify the tester's HGRs and assist the tester in completing the composition using intelligent biosensors. The composition mode is close to the professional composition process.

Awake Surgery for Right Frontal Lobe Glioma: Preserving Emotional Recognition and Facilitating Early Return to Work.

Many glioma patients struggle to return to work after surgery because of higher brain dysfunction. Although the right frontal lobe has historically been considered functionally silent, reports of performing awake surgery to evaluate higher brain functions in patients with tumors in this area have increased. We present two cases of patients who underwent awake surgery for malignant glioma in the right frontal lobe to preserve emotional recognition and facilitate an early return to work. Case 1 was a 48-year-old right-handed woman employed as a nursery school teacher and case 2 was a 21-year-old right-handed man employed in sales. Both had contrast-enhancing right frontal lobe tumors exhibiting high signal intensity on fluid attenuated inversion recovery imaging and underwent awake surgery. During the operation, cortical mapping was performed using the Reading the Mind in the Eyes, calculation, and motor tasks. Resection of sites involved in motor and emotional recognition functions was avoided. In case 1, all regions of high signal intensity were completely resected; in case 2, all regions exhibiting enhancement were resected. Both patients were discharged home without neurological deficits and returned to work within 21 days after surgery. It may be important to focus not only on overall survival and progression-free survival in glioma patients, but also on factors associated with life satisfaction, such as time to return to work after surgery and time until work becomes difficult. Awake surgery aimed at preserving higher brain functions is useful and may also improve life satisfaction.

Chromaticity Recognition Technology of Colored Noise and Operational Modal Analysis

Operational Modal Analysis (OMA) refers to the modal analysis with only output vibration signals of a structure in its operating state. Classic OMA has developed multiple recognition methods in both the time and frequency domains, where when the random excitation is unknown, the excitation chromaticity is usually treated as white color, which can often cause errors and affect the accuracy of identifying frequencies or damping ratios. In this article, the chromaticity recognition function is defined and a method Chromaticity Recognition Technology (CRT) for identifying noise chromaticity based on system response is proposed. Then, a simulation example is presented. The noise chromaticity is identified for the response of the system under four types of colored noise excitation, and the results of the identification of operational mode parameters with and without CRT are compared. Furthermore, the sensitivity of traditional OMA to different colored noise has been investigated. An experiment with a cantilever under base excitation of pink noise has been undertaken and the results demonstrate the feasibility of the proposed CRT in this paper.

Spiking representation learning for associative memories.

Networks of interconnected neurons communicating through spiking signals offer the bedrock of neural computations. Our brain's spiking neural networks have the computational capacity to achieve complex pattern recognition and cognitive functions effortlessly. However, solving real-world problems with artificial spiking neural networks (SNNs) has proved to be difficult for a variety of reasons. Crucially, scaling SNNs to large networks and processing large-scale real-world datasets have been challenging, especially when compared to their non-spiking deep learning counterparts. The critical operation that is needed of SNNs is the ability to learn distributed representations from data and use these representations for perceptual, cognitive and memory operations. In this work, we introduce a novel SNN that performs unsupervised representation learning and associative memory operations leveraging Hebbian synaptic and activity-dependent structural plasticity coupled with neuron-units modelled as Poisson spike generators with sparse firing (~1 Hz mean and ~100 Hz maximum firing rate). Crucially, the architecture of our model derives from the neocortical columnar organization and combines feedforward projections for learning hidden representations and recurrent projections for forming associative memories. We evaluated the model on properties relevant for attractor-based associative memories such as pattern completion, perceptual rivalry, distortion resistance, and prototype extraction.