Knowledge Distillation (KD) facilitates the transfer of discriminative capabilities from an advanced teacher model to a simpler student model, ensuring performance enhancement without compromising accuracy. It is also exploited for model stealing attacks, where adversaries use KD to mimic the functionality of a teacher model. Recent developments in this domain have been influenced by the Stingy Teacher model, which provided empirical analysis showing that sparse outputs can significantly degrade the performance of student models. Addressing the risk of intellectual property leakage, our work introduces an approach to train a teacher model that inherently protects its logits, influenced by the Nasty Teacher concept. Differing from existing methods, we incorporate sparse outputs of adversarial examples with standard training data to strengthen the teacher's defense against student distillation. Our approach carefully reduces the relative entropy between the original and adversarially perturbed outputs, allowing the model to produce adversarial logits with minimal impact on overall performance. The source codes will be made publicly available soon.
This study presents a deep learning-based approach to seismic velocity inversion problem, focusing on both noisy and noiseless training datasets of varying sizes. Our Seismic Velocity Inversion Network (SVInvNet) introduces a novel architecture that contains a multi-connection encoder-decoder structure enhanced with dense blocks. This design is specifically tuned to effectively process complex information, crucial for addressing the challenges of non-linear seismic velocity inversion. For training and testing, we created diverse seismic velocity models, including multi-layered, faulty, and salt dome categories. We also investigated how different kinds of ambient noise, both coherent and stochastic, and the size of the training dataset affect learning outcomes. SVInvNet is trained on datasets ranging from 750 to 6,000 samples and is tested using a large benchmark dataset of 12,000 samples. Despite its fewer parameters compared to the baseline, SVInvNet achieves superior performance with this dataset. The outcomes of the SVInvNet are additionally compared to those of the Full Waveform Inversion (FWI) method. The comparative analysis clearly reveals the effectiveness of the proposed model.
Sign language recognition using computational models is a challenging problem that requires simultaneous spatio-temporal modeling of the multiple sources, i.e. faces, hands, body etc. In this paper, we propose an isolated sign language recognition model based on a model trained using Motion History Images (MHI) that are generated from RGB video frames. RGB-MHI images represent spatio-temporal summary of each sign video effectively in a single RGB image. We propose two different approaches using this model. In the first approach, we use RGB-MHI model as a motion-based spatial attention module integrated in a 3D-CNN architecture. In the second approach, we use RGB-MHI model features directly with a late fusion technique with the features of a 3D-CNN model. We perform extensive experiments on two recently released large-scale isolated sign language datasets, namely AUTSL and BosphorusSign22k datasets. Our experiments show that our models, which use only RGB data, can compete with the state-of-the-art models in the literature that use multi-modal data.
Disease-aware image editing by means of generative adversarial networks (GANs) constitutes a promising avenue for advancing the use of AI in the healthcare sector. Here, we present a proof of concept of this idea. While GAN-based techniques have been successful in generating and manipulating natural images, their application to the medical domain, however, is still in its infancy. Working with the CheXpert data set, we show that StyleGAN can be trained to generate realistic chest X-rays. Inspired by the Cyclic Reverse Generator (CRG) framework, we train an encoder that allows for faithfully inverting the generator on synthetic X-rays and provides organ-level reconstructions of real ones. Employing a guided manipulation of latent codes, we confer the medical condition of cardiomegaly (increased heart size) onto real X-rays from healthy patients. This work was presented in the Medical Imaging meets Neurips Workshop 2020, which was held as part of the 34th Conference on Neural Information Processing Systems (NeurIPS 2020) in Vancouver, Canada
The performances of Sign Language Recognition (SLR) systems have improved considerably in recent years. However, several open challenges still need to be solved to allow SLR to be useful in practice. The research in the field is in its infancy in regards to the robustness of the models to a large diversity of signs and signers, and to fairness of the models to performers from different demographics. This work summarises the ChaLearn LAP Large Scale Signer Independent Isolated SLR Challenge, organised at CVPR 2021 with the goal of overcoming some of the aforementioned challenges. We analyse and discuss the challenge design, top winning solutions and suggestions for future research. The challenge attracted 132 participants in the RGB track and 59 in the RGB+Depth track, receiving more than 1.5K submissions in total. Participants were evaluated using a new large-scale multi-modal Turkish Sign Language (AUTSL) dataset, consisting of 226 sign labels and 36,302 isolated sign video samples performed by 43 different signers. Winning teams achieved more than 96% recognition rate, and their approaches benefited from pose/hand/face estimation, transfer learning, external data, fusion/ensemble of modalities and different strategies to model spatio-temporal information. However, methods still fail to distinguish among very similar signs, in particular those sharing similar hand trajectories.
Remote sensing scene classification deals with the problem of classifying land use/cover of a region from images. To predict the development and socioeconomic structures of cities, the status of land use in regions are tracked by the national mapping agencies of countries. Many of these agencies use land use types that are arranged in multiple levels. In this paper, we examined the efficiency of a hierarchically designed CNN based framework that is suitable for such arrangements. We use NWPU-RESISC45 dataset for our experiments and arranged this data set in a two level nested hierarchy. We have two cascaded deep CNN models initiated using DenseNet-121 architectures. We provide detailed empirical analysis to compare the performances of this hierarchical scheme and its non hierarchical counterpart, together with the individual model performances. We also evaluated the performance of the hierarchical structure statistically to validate the presented empirical results. The results of our experiments show that although individual classifiers for different sub-categories in the hierarchical scheme perform well, the accumulation of classification errors in the cascaded structure prevents its classification performance from exceeding that of the non hierarchical deep model.
Facial image inpainting is a challenging problem as it requires generating new pixels that include semantic information for masked key components in a face, e.g., eyes and nose. Recently, remarkable methods have been proposed in this field. Most of these approaches use encoder-decoder architectures and have different limitations such as allowing unique results for a given image and a particular mask. Alternatively, some approaches generate promising results using different masks with generator networks. However, these approaches are optimization-based and usually require quite a number of iterations. In this paper, we propose an efficient solution to the facial image painting problem using the Cyclic Reverse Generator (CRG) architecture, which provides an encoder-generator model. We use the encoder to embed a given image to the generator space and incrementally inpaint the masked regions until a plausible image is generated; a discriminator network is utilized to assess the generated images during the iterations. We empirically observed that only a few iterations are sufficient to generate realistic images with the proposed model. After the generation process, for the post processing, we utilize a Unet model that we trained specifically for this task to remedy the artifacts close to the mask boundaries. Our method allows applying sketch-based inpaintings, using variety of mask types, and producing multiple and diverse results. We qualitatively compared our method with the state-of-the-art models and observed that our method can compete with the other models in all mask types; it is particularly better in images where larger masks are utilized.
Sign language recognition is a challenging problem where signs are identified by simultaneous local and global articulations of multiple sources, i.e. hand shape and orientation, hand movements, body posture and facial expressions. Solving this problem computationally for a large vocabulary of signs in real life settings is still a challenge, even with the state-of-the-art models. In this study, we present a new large-scale multi-modal Turkish Sign Language dataset (AUTSL) with a benchmark and provide baseline models for performance evaluations. Our dataset consists of 226 signs performed by 43 different signers and 38,336 isolated sign video samples in total. Samples contain a wide variety of backgrounds recorded in indoor and outdoor environments. Moreover, spatial positions and the postures of signers also vary in the recordings. Each sample is recorded with Microsoft Kinect v2 and contains color image (RGB), depth and skeleton data modalities. We prepared benchmark training and test sets for user independent assessments of the models. We trained several deep learning based models and provide empirical evaluations using the benchmark; we used Convolutional Neural Networks (CNNs) to extract features, unidirectional and bidirectional Long Short-Term Memory (LSTM) models to characterize temporal information. We also incorporated feature pooling modules and temporal attention to our models to improve the performances. Using the benchmark test set, we obtained 62.02% accuracy with RGB+Depth data and 47.62% accuracy with RGB only data with the CNN+FPM+BLSTM+Attention model. Our dataset will be made publicly available at https://cvml.ankara.edu.tr.
Isolated sign recognition from video streams is a challenging problem due to the multi-modal nature of the signs, where both local and global hand features and face gestures needs to be attended simultaneously. This problem has recently been studied widely using deep Convolutional Neural Network (CNN) based features and Long Short-Term Memory (LSTM) based deep sequence models. However, the current literature is lack of providing empirical analysis using Hidden Markov Models (HMMs) with deep features. In this study, we provide a framework that is composed of three modules to solve isolated sign recognition problem using different sequence models. The dimensions of deep features are usually too large to work with HMM models. To solve this problem, we propose two alternative CNN based architectures as the second module in our framework, to reduce deep feature dimensions effectively. After extensive experiments, we show that using pretrained Resnet50 features and one of our CNN based dimension reduction models, HMMs can classify isolated signs with 90.15\% accuracy in Montalbano dataset using RGB and Skeletal data. This performance is comparable with the current LSTM based models. HMMs have fewer parameters and can be trained and run on commodity computers fast, without requiring GPUs. Therefore, our analysis with deep features show that HMMs could also be utilized as well as deep sequence models in challenging isolated sign recognition problem.
Image attribute editing is a challenging problem that has been recently studied by many researchers using generative networks. The challenge is in the manipulation of selected attributes of images while preserving the other details. The method to achieve this goal is to find an accurate latent vector representation of an image and a direction corresponding to the attribute. Almost all the works in the literature use labeled datasets in a supervised setting for this purpose. In this study, we introduce an architecture called Cyclic Reverse Generator (CRG), which allows learning the inverse function of the generator accurately via an encoder in an unsupervised setting by utilizing cyclic cost minimization. Attribute editing is then performed using the CRG models for finding desired attribute representations in the latent space. In this work, we use two arbitrary reference images, with and without desired attributes, to compute an attribute direction for editing. We show that the proposed approach performs better in terms of image reconstruction compared to the existing end-to-end generative models both quantitatively and qualitatively. We demonstrate state-of-the-art results on both real images and generated images in MNIST and CelebA datasets.