One-shot Learning Research Articles

A Foreground Prototype-Based One-Shot Segmentation of Brain Tumors.

The potential for enhancing brain tumor segmentation with few-shot learning is enormous. While several deep learning networks (DNNs) show promising segmentation results, they all take a substantial amount of training data in order to yield appropriate results. Moreover, a prominent problem for most of these models is to perform well in unseen classes. To overcome these challenges, we propose a one-shot learning model to segment brain tumors on brain magnetic resonance images (MRI) based on a single prototype similarity score. With the use of recently developed few-shot learning techniques, where training and testing are carried out utilizing support and query sets of images, we attempt to acquire a definitive tumor region by focusing on slices containing foreground classes. It is unlike other recent DNNs that employed the entire set of images. The training of this model is carried out in an iterative manner where in each iteration, random slices containing foreground classes of randomly sampled data are selected as the query set, along with a different random slice from the same sample as the support set. In order to differentiate query images from class prototypes, we used a metric learning-based approach based on non-parametric thresholds. We employed the multimodal Brain Tumor Image Segmentation (BraTS) 2021 dataset with 60 training images and 350 testing images. The effectiveness of the model is evaluated using the mean dice score and mean IoU score. The experimental results provided a dice score of 83.42 which was greater than other works in the literature. Additionally, the proposed one-shot segmentation model outperforms the conventional methods in terms of computational time, memory usage, and the number of data.

Advancing Drug Discovery with Deep Learning: Harnessing Reinforcement Learning and One-Shot Learning for Molecular Design in Low-Data Situations

Drug discovery is a complex process that involves exploring vast chemical spaces to identify potential candidates for the development of effective drugs. While deep learning techniques have shown significant promise in data mining and can be used for molecular design, most drug discovery projects face limitations in low-data situations, making it difficult to train deep learning neural networks. In response to this challenge, this paper proposes a novel drug design system based on deep learning that adopts one-shot learning and reinforcement learning to operate in low-data conditions and generate new molecules with desired properties. Numerical experimental results show that our system can produce valid molecules with desired properties, including high negative logarithm of the half maximal inhibitory concentration (pIC50) values and logarithmic partition co-efficients (log P ) values between 0 and 5. This model is applicable to other molecular design projects with limited data sets, thereby enhancing drug discovery efficiency and effectiveness.

Sleep posture one-shot learning framework based on extremity joint kinematics: In-silico and in-vivo case studies

Sleep posture is linked to several health conditions such as nocturnal cramps and more serious musculoskeletal issues. However, in-clinic sleep assessments are often limited to vital signs (e.g. brain waves). Wearable sensors with embedded inertial measurement units have been used for sleep posture classification; nonetheless, previous works consider only few (commonly four) postures, which are inadequate for advanced clinical assessments. Moreover, posture learning algorithms typically require longitudinal data collection to function reliably, and often operate on raw inertial sensor readings unfamiliar to clinicians. This paper proposes a new framework for sleep posture classification based on a minimal set of joint angle measurements. The proposed framework is validated on a rich set of twelve postures in two experimental pipelines: computer animation to obtain synthetic postural data, and human participant pilot study using custom-made miniature wearable sensors. Through fusing raw geo-inertial sensor measurements to compute a filtered estimate of relative segment orientations across the wrist and ankle joints, the body posture was characterised in a way comprehensible to medical experts. The proposed sleep posture learning framework offers plug-and-play posture classification by capitalising on a novel kinematic data augmentation method that requires only one training example per posture. Additionally, a new metric together with data visualisations were employed to extract meaningful insights from the postures dataset, demonstrate the added value of the data augmentation method, and explain the classification performance. The proposed framework attained promising overall accuracy as high as 100% on synthetic data and 92.7% on real data, on par with state of the art data-hungry algorithms available in the literature.

Dynamic cone-beam CT reconstruction using spatial and temporal implicit neural representation learning (STINR)

Objective. Dynamic cone-beam CT (CBCT) imaging is highly desired in image-guided radiation therapy to provide volumetric images with high spatial and temporal resolutions to enable applications including tumor motion tracking/prediction and intra-delivery dose calculation/accumulation. However, dynamic CBCT reconstruction is a substantially challenging spatiotemporal inverse problem, due to the extremely limited projection sample available for each CBCT reconstruction (one projection for one CBCT volume). Approach. We developed a simultaneous spatial and temporal implicit neural representation (STINR) method for dynamic CBCT reconstruction. STINR mapped the unknown image and the evolution of its motion into spatial and temporal multi-layer perceptrons (MLPs), and iteratively optimized the neuron weightings of the MLPs via acquired projections to represent the dynamic CBCT series. In addition to the MLPs, we also introduced prior knowledge, in the form of principal component analysis (PCA)-based patient-specific motion models, to reduce the complexity of the temporal mapping to address the ill-conditioned dynamic CBCT reconstruction problem. We used the extended-cardiac-torso (XCAT) phantom and a patient 4D-CBCT dataset to simulate different lung motion scenarios to evaluate STINR. The scenarios contain motion variations including motion baseline shifts, motion amplitude/frequency variations, and motion non-periodicity. The XCAT scenarios also contain inter-scan anatomical variations including tumor shrinkage and tumor position change. Main results. STINR shows consistently higher image reconstruction and motion tracking accuracy than a traditional PCA-based method and a polynomial-fitting-based neural representation method. STINR tracks the lung target to an average center-of-mass error of 1–2 mm, with corresponding relative errors of reconstructed dynamic CBCTs around 10%. Significance. STINR offers a general framework allowing accurate dynamic CBCT reconstruction for image-guided radiotherapy. It is a one-shot learning method that does not rely on pre-training and is not susceptible to generalizability issues. It also allows natural super-resolution. It can be readily applied to other imaging modalities as well.

Reservoir Computing for Solving Ordinary Differential Equations

There is a wave of interest in using physics-informed neural networks for solving differential equations. Most of the existing methods are based on feed-forward networks, while recurrent neural networks solvers have not been extensively explored. We introduce a reservoir computing (RC) architecture, an echo-state recurrent neural network capable of discovering approximate solutions that satisfy ordinary differential equations (ODEs). We suggest an approach to calculate time derivatives of recurrent neural network outputs without using back-propagation. The internal weights of an RC are fixed, while only a linear output layer is trained, yielding efficient training. However, RC performance strongly depends on finding the optimal hyper-parameters, which is a computationally expensive process. We use Bayesian optimization to discover optimal sets in a high-dimensional hyper-parameter space efficiently and numerically show that one set is robust and can be transferred to solve an ODE for different initial conditions and time ranges. A closed-form formula for the optimal output weights is derived to solve first-order linear equations in a one-shot backpropagation-free learning process. We extend the RC approach by solving nonlinear systems of ODEs using a hybrid optimization method consisting of gradient descent and Bayesian optimization. Evaluation of linear and nonlinear systems of equations demonstrates the efficiency of the RC ODE solver.

Roof type classification with innovative machine learning approaches.

Recently, convolutional neural network-based methods have been used extensively for roof type classification on images taken from space. The most important problem with classification processes using these methods is that it requires a large amount of training data. Usually, one or a few images are enough for a human to recognise an object. The one-shot learning approach, like the human brain, aims to effect learning about object categories with just one or a few training examples per class, rather than using huge amounts of data. In this study, roof-type classification was carried out with a few training examples using the one-time learning approach and the so-called Siamese neural network method. The images used for training were artificially produced due to the difficulty of finding roof data. A data set consisting of real roof images was used for the test. The test and training data set consisted of three different types: flat, gable and hip. Finally, a convolutional neural network-based model and a Siamese neural network model were trained with the same data set and the test results were compared with each other. When testing the Siamese neural network model, which was trained with artificially produced images, with real roof images, an average classification success of 66% was achieved.

Rapid learning of spatial representations for goal-directed navigation based on a novel model of hippocampal place fields

The discovery of place cells and other spatially modulated neurons in the hippocampal complex of rodents has been crucial to elucidating the neural basis of spatial cognition. More recently, the replay of neural sequences encoding previously experienced trajectories has been observed during consummatory behavior—potentially with implications for rapid learning, quick memory consolidation, and behavioral planning. Several promising models for robotic navigation and reinforcement learning have been proposed based on these and previous findings. Most of these models, however, use carefully engineered neural networks, and sometimes require long learning periods. In this paper, we present a self-organizing model incorporating place cells and replay, and demonstrate its utility for rapid one-shot learning in non-trivial environments with obstacles.

An Intelligent Implementation of Multi-Sensing Data Fusion With Neuromorphic Computing for Human Activity Recognition

The increasing demand for considering multisensor data fusion technology has drawn attention for precise human activity recognition (HAR) over standalone technology due to its reliability and robustness. This article presents a framework that fuses data from multiple sensing systems and applies neuromorphic computing to sense and classify human activities. The data is collected by utilizing inertial measurement unit (IMU) sensors, software-defined radios, and radars, and feature extraction and selection are performed on the data. For each of the actions, such as sitting and standing, an activity matrix is generated, which is then fed into a discrete Hopfield neural network as a binary feature pattern for one-shot learning. Following the Hopfield network neurons’ feedback output, the conformity to the standard activity feature pattern is also determined. Following the Hopfield network neurons’ feedback output, the training of neurons is completed after two steps under the Hebbian learning law, and the conformity to the standard activity feature pattern is also determined. According to the probabilistic statistics on inference predictions, the proposed method, that is the neuromorphic computing of the three data fused framework, achieved the box plot for the highest lower quartile output of 95.34%, while the confusion matrix classification accuracy of the two activities was 98.98%. The results have shown that neuromorphic computing is most capable of multisensor data-fusion-based HAR. Furthermore, the proposed method can be enhanced by incorporating additional hardware signal processing in the system to enable the flexible integration of human activity data.

Similarity function for one-shot learning to enhance the flexibility of myoelectric interfaces.

This study aims to develop a flexible myoelectric pattern recognition (MPR) method based on one-shot learning, which enables convenient switching across different usage scenarios, thereby reducing the re-training burden. First, a one-shot learning model based on a Siamese neural network was constructed to assess the similarity for any given sample pair. In a new scenario involving a new set of gestural categories and/or a new user, just one sample of each category was required to constitute a support set. This enabled the quick deployment of the classifier suitable for the new scenario, which decided for any unknown query sample by selecting the category whose sample in the support set was quantified to be the most like the query sample. The effectiveness of the proposed method was evaluated by experiments conducting MPR across diverse scenarios. The proposed method achieved high recognition accuracy of over 89% under the cross-scenario conditions, and it significantly outperformed other common one-shot learning methods and conventional MPR methods (p < 0.01). This study demonstrates the feasibility of applying one-shot learning to rapidly deploy myoelectric pattern classifiers in response to scenario change. It provides a valuable way of improving the flexibility of myoelectric interfaces toward intelligent gestural control with extensive applications in medical, industrial, and consumer electronics.

Detection of Data Scarce Malware Using One-Shot Learning With Relation Network

Detection of Data Scarce Malware Using One-Shot Learning With Relation Network

Memory-Dependent Computation and Learning in Spiking Neural Networks Through Hebbian Plasticity.

Spiking neural networks (SNNs) are the basis for many energy-efficient neuromorphic hardware systems. While there has been substantial progress in SNN research, artificial SNNs still lack many capabilities of their biological counterparts. In biological neural systems, memory is a key component that enables the retention of information over a huge range of temporal scales, ranging from hundreds of milliseconds up to years. While Hebbian plasticity is believed to play a pivotal role in biological memory, it has so far been analyzed mostly in the context of pattern completion and unsupervised learning in artificial and SNNs. Here, we propose that Hebbian plasticity is fundamental for computations in biological and artificial spiking neural systems. We introduce a novel memory-augmented SNN architecture that is enriched by Hebbian synaptic plasticity. We show that Hebbian enrichment renders SNNs surprisingly versatile in terms of their computational as well as learning capabilities. It improves their abilities for out-of-distribution generalization, one-shot learning, cross-modal generative association, language processing, and reward-based learning. This suggests that powerful cognitive neuromorphic systems can be built based on this principle.

DSW: One-shot Learning Scheme for Device-free Acoustic Gesture Signals

DSW: One-shot Learning Scheme for Device-free Acoustic Gesture Signals

One-Shot Learning for Periocular Recognition: Exploring the Effect of Domain Adaptation and Data Bias on Deep Representations

One-Shot Learning for Periocular Recognition: Exploring the Effect of Domain Adaptation and Data Bias on Deep Representations

An Extended Spatiotemporal Contextual Learning and Memory Network Model for Hardware Implementation

A spatiotemporal contextual learning and memory network (STCLMN) model was proposed based on the hippocampal CA1 network using the difference equations suitable for digital computer simulations. However, the inherent capability of handling and processing of continuous-time asynchronous signals (spikes) cannot be fully employed in the STCLMN model. Furthermore, digital circuit implementation of the STCLMN is not suitable for real-time edge applications, wherein small size, low power consumption, and real-time learning capability are necessary. Therefore, this study extends the original STCLMN model using a set of differential equations to overcome the above problems encountered by STCLMN for asynchronous circuit implementation. The minimum accuracy required for a compact and low-power hardware implementation of the proposed model is then investigated. The validity and feasibility of the proposed model are confirmed through numerical simulations and experiments with a small mixed analog/digital circuit based on the investigated accuracy. In this study, LIF neurons are used because of their ease of circuit implementation. The proposed eSTCLMN model would enable an efficient circuit implementation of an asynchronous spiking neural network with spatiotemporal contextual one-shot learning and memory.

Electromagnetic Imaging Boosted Visual Object Recognition Under Difficult Visual Conditions

Object imaging and recognition under difficult visual conditions is extremely challenging due to the captured low-quality images, and traditional optical-based recognition methods always fail in this task. In this paper, we propose to utilize the visual-microwave image pairs captured by both visual cameras and microwave sensors for imaging and recognition. To address the heavy noises in the low-quality optical images, we retrieve the physically quantitative images from associated scattered field data, and enhance visual features by both optical and retrieval images. We develop a cross-modal Enhanced Attentive Visual-Microwave Fusion (EAVMF) object recognition model to jointly learn the cross-modal generator and multimodal recognizer. In addition, an attention module for the visual subnetwork is utilized to highlight the regions of interest. Two multimodal datasets with synthetic visual-microwave image pairs are built to simulate the difficult visual condition. The numerical results on these datasets demonstrate that: 1) both the multimodal fusion, cross-modal enhancement, and visual attention module can enhance the performance; and 2) compared with existing methods, the proposed EAVMF not only performs better in terms of accuracy but also has good scalability and one-shot learning ability.

Multi-spectral template matching based object detection in a few-shot learning manner

Multi-spectral template matching (MSTM) based object detection approaches can be widely used in robotics and aerospace systems for fine-grained object discovery. However, the performance of existing nearest neighbor search based nonparametric paradigms (e.g., correlation coefficient and lp-norm) turns out to be unsatisfactory. These paradigms tend to suffer from two defects: 1) they fail to select from the raw features the discriminative ones that can help distinguish between the target and background; 2) the domain shift between the template and search spectra has not been well addressed within the feature space. In this work, we propose a data-driven MSTM method to address these two issues. First, Exemplar-SVM (E-SVM) is applied to execute feature selection and target/background categorization jointly, which is facilitated by its max-margin mechanism. To enable the learning process where the template is regarded as a single positive sample, knowledge transfer is executed to attain negative samples from other domains, e.g., large-scale public datasets. Then, the hard negative samples are mined to help train a discriminative classifier. Concerning practical applications, we also augment the template with different image degradations and extend E-SVM from the original one-shot learning approach to its few-shot version. Second, a multi-domain adaptation approach via unsupervised multi-domain subspace alignment is proposed to tackle multi-domain shift problem. Here the multiple domains relate to template, search, and negative ones considering both offline learning and online matching. The wide-range experimental results on two multi-spectral datasets demonstrate the effectiveness of our method. The tailored dataset and code will be released publicly.

Robotic Navigation Based on Experiences and Predictive Map Inspired by Spatial Cognition

Humans and animals have environmental cognition and navigation abilities. These abilities are closely related to the spatial cognitive mechanism of brain. Based on this mechanism, we propose a novel robotic navigation framework based on experiences and predictive map inspired by spatial cognition to accurately construct environment experience and quickly plan path. The grid cell and the place cell are modeled to rapidly integrate self-motion cues. The multidimensional grid coding and one-shot learning rule are utilized for activating the place representation of robot pose. Visual cues provide information for relocation. These information are integrated through experiences which express the topology of the environment, enabling the robot to accurately achieve spatial cognition of complex environment. In order to realize the sequential decision making of hippocampus, the predictive map is introduced to quickly plan the experience sequence to target in dynamic environment. The successor representation model of robot's state is constructed through reinforcement learning. Combined with the goal-based reward function, the shortest path can be replanned to adapt to environment changes. The proposed method is tested in simulated maze, Kitti dataset and corridor environment. Compared with other bionic navigation methods, it has a faster computing speed and higher precision with bionic characteristics. Compared with visual navigation methods, the navigation tasks can robustly be accomplished without complex system design.

Brain-inspired Cognition in Next-generation Racetrack Memories

Hyperdimensional computing (HDC) is an emerging computational framework inspired by the brain that operates on vectors with thousands of dimensions to emulate cognition. Unlike conventional computational frameworks that operate on numbers, HDC, like the brain, uses high-dimensional random vectors and is capable of one-shot learning. HDC is based on a well-defined set of arithmetic operations and is highly error resilient. The core operations of HDC manipulate HD vectors in bulk bit-wise fashion, offering many opportunities to leverage parallelism. Unfortunately, on conventional von Neumann architectures, the continuous movement of HD vectors among the processor and the memory can make the cognition task prohibitively slow and energy intensive. Hardware accelerators only marginally improve related metrics. In contrast, even partial implementations of an HDC framework inside memory can provide considerable performance/energy gains as demonstrated in prior work using memristors. This article presents an architecture based on racetrack memory (RTM) to conduct and accelerate the entire HDC framework within memory. The proposed solution requires minimal additional CMOS circuitry by leveraging a read operation across multiple domains in RTMs called transverse read (TR) to realize exclusive-or ( XOR ) and addition operations. To minimize the CMOS circuitry overhead, an RTM nanowire-based counting mechanism is proposed. Using language recognition as the example workload, the proposed RTM HDC system reduces the energy consumption by 8.6× compared to the state-of-the-art in-memory implementation. Compared to dedicated hardware design realized with an FPGA, RTM-based HDC processing demonstrates 7.8× and 5.3× improvements in the overall runtime and energy consumption, respectively.

One-shot learning of stochastic differential equations with data adapted kernels

We consider the problem of learning Stochastic Differential Equations of the form dXt=f(Xt)dt+σ(Xt)dWt from one sample trajectory. This problem is more challenging than learning deterministic dynamical systems because one sample trajectory only provides indirect information on the unknown functions f, σ, and stochastic process dWt representing the drift, the diffusion, and the stochastic forcing terms, respectively. We propose a method that combines Computational Graph Completion [1] and data adapted kernels learned via a new variant of cross validation. Our approach can be decomposed as follows: (1) Represent the time-increment map Xt→Xt+dt as a Computational Graph in which f, σ and dWt appear as unknown functions and random variables. (2) Complete the graph (approximate unknown functions and random variables) via Maximum a Posteriori Estimation (given the data) with Gaussian Process (GP) priors on the unknown functions. (3) Learn the covariance functions (kernels) of the GP priors from data with randomized cross-validation. Numerical experiments illustrate the efficacy, robustness, and scope of our method.

Leveraging siamese networks for one-shot intrusion detection model

The use of supervised Machine Learning (ML) to enhance Intrusion Detection Systems (IDS) has been the subject of significant research. Supervised ML is based upon learning by example, demanding significant volumes of representative instances for effective training and the need to retrain the model for every unseen cyber-attack class. However, retraining the models in-situ renders the network susceptible to attacks owing to the time-window required to acquire a sufficient volume of data. Although anomaly detection systems provide a coarse-grained defence against unseen attacks, these approaches are significantly less accurate and suffer from high false-positive rates. Here, a complementary approach referred to as “One-Shot Learning”, whereby a limited number of examples of a new attack-class is used to identify a new attack-class (out of many) is detailed. The model grants a new cyber-attack classification opportunity for classes that were not seen during training without retraining. A Siamese Network is trained to differentiate between classes based on pairs similarities, rather than features, allowing to identify new and previously unseen attacks. The performance of a pre-trained model to classify new attack-classes based only on one example is evaluated using three mainstream IDS datasets; CICIDS2017, NSL-KDD, and KDD Cup’99. The results confirm the adaptability of the model in classifying unseen attacks and the trade-off between performance and the need for distinctive class representations.