arXiv subjects

Radu Tudor Ionescu

Publications and source records attributed to Radu Tudor Ionescu.

At least 91 records · Page 5Linked to original sources

arXiv2022

Feature-level augmentation to improve robustness of deep neural networks to affine transformations

Recent studies revealed that convolutional neural networks do not generalize well to small image transformations, e.g. rotations by a few degrees or translations of a few pixels. To improve the robustness to such transformations, we propose to introduce data augmentation at intermediate layers of the neural architecture, in addition to the common data augmentation applied on the input images. By introducing small perturbations to activation maps (features) at various levels, we develop the capacity of the neural network to cope with such transformations. We conduct experiments on three image classification benchmarks (Tiny ImageNet, Caltech-256 and Food-101), considering two different convolutional architectures (ResNet-18 and DenseNet-121). When compared with two state-of-the-art stabilization methods, the empirical results show that our approach consistently attains the best trade-off between accuracy and mean flip rate.

cs.CV↗

arXiv2022

Self-paced learning to improve text row detection in historical documents with missing labels

An important preliminary step of optical character recognition systems is the detection of text rows. To address this task in the context of historical data with missing labels, we propose a self-paced learning algorithm capable of improving the row detection performance. We conjecture that pages with more ground-truth bounding boxes are less likely to have missing annotations. Based on this hypothesis, we sort the training examples in descending order with respect to the number of ground-truth bounding boxes, and organize them into k batches. Using our self-paced learning method, we train a row detector over k iterations, progressively adding batches with less ground-truth annotations. At each iteration, we combine the ground-truth bounding boxes with pseudo-bounding boxes (bounding boxes predicted by the model itself) using non-maximum suppression, and we include the resulting annotations at the next training iteration. We demonstrate that our self-paced learning strategy brings significant performance gains on two data sets of historical documents, improving the average precision of YOLOv4 with more than 12% on one data set and 39% on the other.

cs.CV↗

arXiv2022

SepTr: Separable Transformer for Audio Spectrogram Processing

Following the successful application of vision transformers in multiple computer vision tasks, these models have drawn the attention of the signal processing community. This is because signals are often represented as spectrograms (e.g. through Discrete Fourier Transform) which can be directly provided as input to vision transformers. However, naively applying transformers to spectrograms is suboptimal. Since the axes represent distinct dimensions, i.e. frequency and time, we argue that a better approach is to separate the attention dedicated to each axis. To this end, we propose the Separable Transformer (SepTr), an architecture that employs two transformer blocks in a sequential manner, the first attending to tokens within the same time interval, and the second attending to tokens within the same frequency bin. We conduct experiments on three benchmark data sets, showing that our separable architecture outperforms conventional vision transformers and other state-of-the-art methods. Unlike standard transformers, SepTr linearly scales the number of trainable parameters with the input size, thus having a lower memory footprint. Our code is available as open source at https://github.com/ristea/septr.

cs.CV↗

arXiv2022

Curriculum Learning: A Survey

Training machine learning models in a meaningful order, from the easy samples to the hard ones, using curriculum learning can provide performance improvements over the standard training approach based on random data shuffling, without any additional computational costs. Curriculum learning strategies have been successfully employed in all areas of machine learning, in a wide range of tasks. However, the necessity of finding a way to rank the samples from easy to hard, as well as the right pacing function for introducing more difficult data can limit the usage of the curriculum approaches. In this survey, we show how these limits have been tackled in the literature, and we present different curriculum learning instantiations for various tasks in machine learning. We construct a multi-perspective taxonomy of curriculum learning approaches by hand, considering various classification criteria. We further build a hierarchical tree of curriculum learning methods using an agglomerative clustering algorithm, linking the discovered clusters with our taxonomy. At the end, we provide some interesting directions for future work.

cs.LG↗

arXiv2022

Discriminability-enforcing loss to improve representation learning

During the training process, deep neural networks implicitly learn to represent the input data samples through a hierarchy of features, where the size of the hierarchy is determined by the number of layers. In this paper, we focus on enforcing the discriminative power of the high-level representations, that are typically learned by the deeper layers (closer to the output). To this end, we introduce a new loss term inspired by the Gini impurity, which is aimed at minimizing the entropy (increasing the discriminative power) of individual high-level features with respect to the class labels. Although our Gini loss induces highly-discriminative features, it does not ensure that the distribution of the high-level features matches the distribution of the classes. As such, we introduce another loss term to minimize the Kullback-Leibler divergence between the two distributions. We conduct experiments on two image classification data sets (CIFAR-100 and Caltech 101), considering multiple neural architectures ranging from convolutional networks (ResNet-17, ResNet-18, ResNet-50) to transformers (CvT). Our empirical results show that integrating our novel loss terms into the training objective consistently outperforms the models trained with cross-entropy alone, without increasing the inference time at all.

cs.CV↗

arXiv2022

Self-Supervised Predictive Convolutional Attentive Block for Anomaly Detection

Anomaly detection is commonly pursued as a one-class classification problem, where models can only learn from normal training samples, while being evaluated on both normal and abnormal test samples. Among the successful approaches for anomaly detection, a distinguished category of methods relies on predicting masked information (e.g. patches, future frames, etc.) and leveraging the reconstruction error with respect to the masked information as an abnormality score. Different from related methods, we propose to integrate the reconstruction-based functionality into a novel self-supervised predictive architectural building block. The proposed self-supervised block is generic and can easily be incorporated into various state-of-the-art anomaly detection methods. Our block starts with a convolutional layer with dilated filters, where the center area of the receptive field is masked. The resulting activation maps are passed through a channel attention module. Our block is equipped with a loss that minimizes the reconstruction error with respect to the masked area in the receptive field. We demonstrate the generality of our block by integrating it into several state-of-the-art frameworks for anomaly detection on image and video, providing empirical evidence that shows considerable performance improvements on MVTec AD, Avenue, and ShanghaiTech. We release our code as open source at https://github.com/ristea/sspcab.

cs.CV↗

arXiv2021

Teacher-Student Training and Triplet Loss to Reduce the Effect of Drastic Face Occlusion

We study a series of recognition tasks in two realistic scenarios requiring the analysis of faces under strong occlusion. On the one hand, we aim to recognize facial expressions of people wearing Virtual Reality (VR) headsets. On the other hand, we aim to estimate the age and identify the gender of people wearing surgical masks. For all these tasks, the common ground is that half of the face is occluded. In this challenging setting, we show that convolutional neural networks (CNNs) trained on fully-visible faces exhibit very low performance levels. While fine-tuning the deep learning models on occluded faces is extremely useful, we show that additional performance gains can be obtained by distilling knowledge from models trained on fully-visible faces. To this end, we study two knowledge distillation methods, one based on teacher-student training and one based on triplet loss. Our main contribution consists in a novel approach for knowledge distillation based on triplet loss, which generalizes across models and tasks. Furthermore, we consider combining distilled models learned through conventional teacher-student training or through our novel teacher-student training based on triplet loss. We provide empirical evidence showing that, in most cases, both individual and combined knowledge distillation methods bring statistically significant performance improvements. We conduct experiments with three different neural models (VGG-f, VGG-face, ResNet-50) on various tasks (facial expression recognition, gender recognition, age estimation), showing consistent improvements regardless of the model or task.

cs.CV↗

arXiv2021

The Unreasonable Effectiveness of Machine Learning in Moldavian versus Romanian Dialect Identification

Motivated by the seemingly high accuracy levels of machine learning models in Moldavian versus Romanian dialect identification and the increasing research interest on this topic, we provide a follow-up on the Moldavian versus Romanian Cross-Dialect Topic Identification (MRC) shared task of the VarDial 2019 Evaluation Campaign. The shared task included two sub-task types: one that consisted in discriminating between the Moldavian and Romanian dialects and one that consisted in classifying documents by topic across the two dialects of Romanian. Participants achieved impressive scores, e.g. the top model for Moldavian versus Romanian dialect identification obtained a macro F1 score of 0.895. We conduct a subjective evaluation by human annotators, showing that humans attain much lower accuracy rates compared to machine learning (ML) models. Hence, it remains unclear why the methods proposed by participants attain such high accuracy rates. Our goal is to understand (i) why the proposed methods work so well (by visualizing the discriminative features) and (ii) to what extent these methods can keep their high accuracy levels, e.g. when we shorten the text samples to single sentences or when we use tweets at inference time. A secondary goal of our work is to propose an improved ML model using ensemble learning. Our experiments show that ML models can accurately identify the dialects, even at the sentence level and across different domains (news articles versus tweets). We also analyze the most discriminative features of the best performing models, providing some explanations behind the decisions taken by these models. Interestingly, we learn new dialectal patterns previously unknown to us or to our human annotators. Furthermore, we conduct experiments showing that the machine learning performance on the MRC shared task can be improved through an ensemble based on stacking.

cs.CL↗

arXiv2021

Estimating the Magnitude and Phase of Automotive Radar Signals under Multiple Interference Sources with Fully Convolutional Networks

Radar sensors are gradually becoming a wide-spread equipment for road vehicles, playing a crucial role in autonomous driving and road safety. The broad adoption of radar sensors increases the chance of interference among sensors from different vehicles, generating corrupted range profiles and range-Doppler maps. In order to extract distance and velocity of multiple targets from range-Doppler maps, the interference affecting each range profile needs to be mitigated. In this paper, we propose a fully convolutional neural network for automotive radar interference mitigation. In order to train our network in a real-world scenario, we introduce a new data set of realistic automotive radar signals with multiple targets and multiple interferers. To our knowledge, we are the first to apply weight pruning in the automotive radar domain, obtaining superior results compared to the widely-used dropout. While most previous works successfully estimated the magnitude of automotive radar signals, we propose a deep learning model that can accurately estimate the phase. For instance, our novel approach reduces the phase estimation error with respect to the commonly-adopted zeroing technique by half, from 12.55 degrees to 6.58 degrees. Considering the lack of databases for automotive radar interference mitigation, we release as open source our large-scale data set that closely replicates the real-world automotive scenario for multiple interference cases, allowing others to objectively compare their future work in this domain. Our data set is available for download at: http://github.com/ristea/arim-v2.

eess.SP↗

arXiv2021

A realistic approach to generate masked faces applied on two novel masked face recognition data sets

The COVID-19 pandemic raises the problem of adapting face recognition systems to the new reality, where people may wear surgical masks to cover their noses and mouths. Traditional data sets (e.g., CelebA, CASIA-WebFace) used for training these systems were released before the pandemic, so they now seem unsuited due to the lack of examples of people wearing masks. We propose a method for enhancing data sets containing faces without masks by creating synthetic masks and overlaying them on faces in the original images. Our method relies on SparkAR Studio, a developer program made by Facebook that is used to create Instagram face filters. In our approach, we use 9 masks of different colors, shapes and fabrics. We employ our method to generate a number of 445,446 (90%) samples of masks for the CASIA-WebFace data set and 196,254 (96.8%) masks for the CelebA data set, releasing the mask images at https://github.com/securifai/masked_faces. We show that our method produces significantly more realistic training examples of masks overlaid on faces by asking volunteers to qualitatively compare it to other methods or data sets designed for the same task. We also demonstrate the usefulness of our method by evaluating state-of-the-art face recognition systems (FaceNet, VGG-face, ArcFace) trained on our enhanced data sets and showing that they outperform equivalent systems trained on original data sets (containing faces without masks) or competing data sets (containing masks generated by related methods), when the test benchmarks contain masked faces.

cs.CV↗

arXiv2021

Anomaly Detection in Video via Self-Supervised and Multi-Task Learning

Anomaly detection in video is a challenging computer vision problem. Due to the lack of anomalous events at training time, anomaly detection requires the design of learning methods without full supervision. In this paper, we approach anomalous event detection in video through self-supervised and multi-task learning at the object level. We first utilize a pre-trained detector to detect objects. Then, we train a 3D convolutional neural network to produce discriminative anomaly-specific information by jointly learning multiple proxy tasks: three self-supervised and one based on knowledge distillation. The self-supervised tasks are: (i) discrimination of forward/backward moving objects (arrow of time), (ii) discrimination of objects in consecutive/intermittent frames (motion irregularity) and (iii) reconstruction of object-specific appearance information. The knowledge distillation task takes into account both classification and detection information, generating large prediction discrepancies between teacher and student models when anomalies occur. To the best of our knowledge, we are the first to approach anomalous event detection in video as a multi-task learning problem, integrating multiple self-supervised and knowledge distillation proxy tasks in a single architecture. Our lightweight architecture outperforms the state-of-the-art methods on three benchmarks: Avenue, ShanghaiTech and UCSD Ped2. Additionally, we perform an ablation study demonstrating the importance of integrating self-supervised learning and normality-specific distillation in a multi-task learning setting.

cs.CV↗

arXiv2021

Contextual Convolutional Neural Networks

We propose contextual convolution (CoConv) for visual recognition. CoConv is a direct replacement of the standard convolution, which is the core component of convolutional neural networks. CoConv is implicitly equipped with the capability of incorporating contextual information while maintaining a similar number of parameters and computational cost compared to the standard convolution. CoConv is inspired by neuroscience studies indicating that (i) neurons, even from the primary visual cortex (V1 area), are involved in detection of contextual cues and that (ii) the activity of a visual neuron can be influenced by the stimuli placed entirely outside of its theoretical receptive field. On the one hand, we integrate CoConv in the widely-used residual networks and show improved recognition performance over baselines on the core tasks and benchmarks for visual recognition, namely image classification on the ImageNet data set and object detection on the MS COCO data set. On the other hand, we introduce CoConv in the generator of a state-of-the-art Generative Adversarial Network, showing improved generative results on CIFAR-10 and CelebA. Our code is available at https://github.com/iduta/coconv.

cs.CV↗

arXiv2021

Improving the Authentication with Built-in Camera Protocol Using Built-in Motion Sensors: A Deep Learning Solution

We propose an enhanced version of the Authentication with Built-in Camera (ABC) protocol by employing a deep learning solution based on built-in motion sensors. The standard ABC protocol identifies mobile devices based on the photo-response non-uniformity (PRNU) of the camera sensor, while also considering QR-code-based meta-information. During authentication, the user is required to take two photos that contain two QR codes presented on a screen. The presented QR code images also contain a unique probe signal, similar to a camera fingerprint, generated by the protocol. During verification, the server computes the fingerprint of the received photos and authenticates the user if (i) the probe signal is present, (ii) the metadata embedded in the QR codes is correct and (iii) the camera fingerprint is identified correctly. However, the protocol is vulnerable to forgery attacks when the attacker can compute the camera fingerprint from external photos, as shown in our preliminary work. In this context, we propose an enhancement for the ABC protocol based on motion sensor data, as an additional and passive authentication layer. Smartphones can be identified through their motion sensor data, which, unlike photos, is never posted by users on social media platforms, thus being more secure than using photographs alone. To this end, we transform motion signals into embedding vectors produced by deep neural networks, applying Support Vector Machines for the smartphone identification task. Our change to the ABC protocol results in a multi-modal protocol that lowers the false acceptance rate for the attack proposed in our previous work to a percentage as low as 0.07%.

cs.CR↗

arXiv2021

SaRoCo: Detecting Satire in a Novel Romanian Corpus of News Articles

In this work, we introduce a corpus for satire detection in Romanian news. We gathered 55,608 public news articles from multiple real and satirical news sources, composing one of the largest corpora for satire detection regardless of language and the only one for the Romanian language. We provide an official split of the text samples, such that training news articles belong to different sources than test news articles, thus ensuring that models do not achieve high performance simply due to overfitting. We conduct experiments with two state-of-the-art deep neural models, resulting in a set of strong baselines for our novel corpus. Our results show that the machine-level accuracy for satire detection in Romanian is quite low (under 73% on the test set) compared to the human-level accuracy (87%), leaving enough room for improvement in future research.

cs.CL↗

arXiv2021

Unsupervised Medical Image Alignment with Curriculum Learning

We explore different curriculum learning methods for training convolutional neural networks on the task of deformable pairwise 3D medical image registration. To the best of our knowledge, we are the first to attempt to improve performance by training medical image registration models using curriculum learning, starting from an easy training setup in the first training stages, and gradually increasing the complexity of the setup. On the one hand, we consider two existing curriculum learning approaches, namely curriculum dropout and curriculum by smoothing. On the other hand, we propose a novel and simple strategy to achieve curriculum, namely to use purposely blurred images at the beginning, then gradually transit to sharper images in the later training stages. Our experiments with an underlying state-of-the-art deep learning model show that curriculum learning can lead to superior results compared to conventional training. Additionally, we show that curriculum by input blur has the best accuracy versus speed trade-off among the compared curriculum learning approaches.

cs.CV↗

arXiv2021

Self-paced ensemble learning for speech and audio classification

Combining multiple machine learning models into an ensemble is known to provide superior performance levels compared to the individual components forming the ensemble. This is because models can complement each other in taking better decisions. Instead of just combining the models, we propose a self-paced ensemble learning scheme in which models learn from each other over several iterations. During the self-paced learning process based on pseudo-labeling, in addition to improving the individual models, our ensemble also gains knowledge about the target domain. To demonstrate the generality of our self-paced ensemble learning (SPEL) scheme, we conduct experiments on three audio tasks. Our empirical results indicate that SPEL significantly outperforms the baseline ensemble models. We also show that applying self-paced learning on individual models is less effective, illustrating the idea that models in the ensemble actually learn from each other.

cs.SD↗

arXiv2021

FreSaDa: A French Satire Data Set for Cross-Domain Satire Detection

In this paper, we introduce FreSaDa, a French Satire Data Set, which is composed of 11,570 articles from the news domain. In order to avoid reporting unreasonably high accuracy rates due to the learning of characteristics specific to publication sources, we divided our samples into training, validation and test, such that the training publication sources are distinct from the validation and test publication sources. This gives rise to a cross-domain (cross-source) satire detection task. We employ two classification methods as baselines for our new data set, one based on low-level features (character n-grams) and one based on high-level features (average of CamemBERT word embeddings). As an additional contribution, we present an unsupervised domain adaptation method based on regarding the pairwise similarities (given by the dot product) between the training samples and the validation samples as features. By including these domain-specific features, we attain significant improvements for both character n-grams and CamemBERT embeddings.

cs.CL↗

arXiv2021

Automotive Radar Interference Mitigation with Unfolded Robust PCA based on Residual Overcomplete Auto-Encoder Blocks

In autonomous driving, radar systems play an important role in detecting targets such as other vehicles on the road. Radars mounted on different cars can interfere with each other, degrading the detection performance. Deep learning methods for automotive radar interference mitigation can succesfully estimate the amplitude of targets, but fail to recover the phase of the respective targets. In this paper, we propose an efficient and effective technique based on unfolded robust Principal Component Analysis (RPCA) that is able to estimate both amplitude and phase in the presence of interference. Our contribution consists in introducing residual overcomplete auto-encoder (ROC-AE) blocks into the recurrent architecture of unfolded RPCA, which results in a deeper model that significantly outperforms unfolded RPCA as well as other deep learning models.

eess.SP↗