FeSViBS: Federated Split Learning of Vision Transformer with Block Sampling

Data scarcity is a significant obstacle hindering the learning of powerful machine learning models in critical healthcare applications. Data-sharing mechanisms among multiple entities (e.g., hospitals) can accelerate model training and yield more accurate predictions. Recently, approaches such as Federated Learning (FL) and Split Learning (SL) have facilitated collaboration without the need to exchange private data. In this work, we propose a framework for medical imaging classification tasks called Federated Split learning of Vision transformer with Block Sampling (FeSViBS). The FeSViBS framework builds upon the existing federated split vision transformer and introduces a block sampling module, which leverages intermediate features extracted by the Vision Transformer (ViT) at the server. This is achieved by sampling features (patch tokens) from an intermediate transformer block and distilling their information content into a pseudo class token before passing them back to the client. These pseudo class tokens serve as an effective feature augmentation strategy and enhances the generalizability of the learned model. We demonstrate the utility of our proposed method compared to other SL and FL approaches on three publicly available medical imaging datasets: HAM1000, BloodMNIST, and Fed-ISIC2019, under both IID and non-IID settings. Code: https://github.com/faresmalik/FeSViBS

Evading Forensic Classifiers with Attribute-Conditioned Adversarial Faces

The ability of generative models to produce highly realistic synthetic face images has raised security and ethical concerns. As a first line of defense against such fake faces, deep learning based forensic classifiers have been developed. While these forensic models can detect whether a face image is synthetic or real with high accuracy, they are also vulnerable to adversarial attacks. Although such attacks can be highly successful in evading detection by forensic classifiers, they introduce visible noise patterns that are detectable through careful human scrutiny. Additionally, these attacks assume access to the target model(s) which may not always be true. Attempts have been made to directly perturb the latent space of GANs to produce adversarial fake faces that can circumvent forensic classifiers. In this work, we go one step further and show that it is possible to successfully generate adversarial fake faces with a specified set of attributes (e.g., hair color, eye size, race, gender, etc.). To achieve this goal, we leverage the state-of-the-art generative model StyleGAN with disentangled representations, which enables a range of modifications without leaving the manifold of natural images. We propose a framework to search for adversarial latent codes within the feature space of StyleGAN, where the search can be guided either by a text prompt or a reference image. We also propose a meta-learning based optimization strategy to achieve transferable performance on unknown target models. Extensive experiments demonstrate that the proposed approach can produce semantically manipulated adversarial fake faces, which are true to the specified attribute set and can successfully fool forensic face classifiers, while remaining undetectable by humans. Code: https://github.com/koushiksrivats/face_attribute_attack.

CLIP2Protect: Protecting Facial Privacy using Text-Guided Makeup via Adversarial Latent Search

The success of deep learning based face recognition systems has given rise to serious privacy concerns due to their ability to enable unauthorized tracking of users in the digital world. Existing methods for enhancing privacy fail to generate naturalistic images that can protect facial privacy without compromising user experience. We propose a novel two-step approach for facial privacy protection that relies on finding adversarial latent codes in the low-dimensional manifold of a pretrained generative model. The first step inverts the given face image into the latent space and finetunes the generative model to achieve an accurate reconstruction of the given image from its latent code. This step produces a good initialization, aiding the generation of high-quality faces that resemble the given identity. Subsequently, user-defined makeup text prompts and identity-preserving regularization are used to guide the search for adversarial codes in the latent space. Extensive experiments demonstrate that faces generated by our approach have stronger black-box transferability with an absolute gain of 12.06% over the state-of-the-art facial privacy protection approach under the face verification task. Finally, we demonstrate the effectiveness of the proposed approach for commercial face recognition systems. Our code is available at https://github.com/fahadshamshad/Clip2Protect.

Single-branch Network for Multimodal Training

With the rapid growth of social media platforms, users are sharing billions of multimedia posts containing audio, images, and text. Researchers have focused on building autonomous systems capable of processing such multimedia data to solve challenging multimodal tasks including cross-modal retrieval, matching, and verification. Existing works use separate networks to extract embeddings of each modality to bridge the gap between them. The modular structure of their branched networks is fundamental in creating numerous multimodal applications and has become a defacto standard to handle multiple modalities. In contrast, we propose a novel single-branch network capable of learning discriminative representation of unimodal as well as multimodal tasks without changing the network. An important feature of our single-branch network is that it can be trained either using single or multiple modalities without sacrificing performance. We evaluated our proposed single-branch network on the challenging multimodal problem (face-voice association) for cross-modal verification and matching tasks with various loss formulations. Experimental results demonstrate the superiority of our proposed single-branch network over the existing methods in a wide range of experiments. Code: https://github.com/msaadsaeed/SBNet

On the Importance of Image Encoding in Automated Chest X-Ray Report Generation

Chest X-ray is one of the most popular medical imaging modalities due to its accessibility and effectiveness. However, there is a chronic shortage of well-trained radiologists who can interpret these images and diagnose the patient's condition. Therefore, automated radiology report generation can be a very helpful tool in clinical practice. A typical report generation workflow consists of two main steps: (i) encoding the image into a latent space and (ii) generating the text of the report based on the latent image embedding. Many existing report generation techniques use a standard convolutional neural network (CNN) architecture for image encoding followed by a Transformer-based decoder for medical text generation. In most cases, CNN and the decoder are trained jointly in an end-to-end fashion. In this work, we primarily focus on understanding the relative importance of encoder and decoder components. Towards this end, we analyze four different image encoding approaches: direct, fine-grained, CLIP-based, and Cluster-CLIP-based encodings in conjunction with three different decoders on the large-scale MIMIC-CXR dataset. Among these encoders, the cluster CLIP visual encoder is a novel approach that aims to generate more discriminative and explainable representations. CLIP-based encoders produce comparable results to traditional CNN-based encoders in terms of NLP metrics, while fine-grained encoding outperforms all other encoders both in terms of NLP and clinical accuracy metrics, thereby validating the importance of image encoder to effectively extract semantic information. GitHub repository: https://github.com/mudabek/encoding-cxr-report-gen

Towards Building Text-To-Speech Systems for the Next Billion Users

Deep learning based text-to-speech (TTS) systems have been evolving rapidly with advances in model architectures, training methodologies, and generalization across speakers and languages. However, these advances have not been thoroughly investigated for Indian language speech synthesis. Such investigation is computationally expensive given the number and diversity of Indian languages, relatively lower resource availability, and the diverse set of advances in neural TTS that remain untested. In this paper, we evaluate the choice of acoustic models, vocoders, supplementary loss functions, training schedules, and speaker and language diversity for Dravidian and Indo-Aryan languages. Based on this, we identify monolingual models with FastPitch and HiFi-GAN V1, trained jointly on male and female speakers to perform the best. With this setup, we train and evaluate TTS models for 13 languages and find our models to significantly improve upon existing models in all languages as measured by mean opinion scores. We open-source all models on the Bhashini platform.

Hate-CLIPper: Multimodal Hateful Meme Classification based on Cross-modal Interaction of CLIP Features

Hateful memes are a growing menace on social media. While the image and its corresponding text in a meme are related, they do not necessarily convey the same meaning when viewed individually. Hence, detecting hateful memes requires careful consideration of both visual and textual information. Multimodal pre-training can be beneficial for this task because it effectively captures the relationship between the image and the text by representing them in a similar feature space. Furthermore, it is essential to model the interactions between the image and text features through intermediate fusion. Most existing methods either employ multimodal pre-training or intermediate fusion, but not both. In this work, we propose the Hate-CLIPper architecture, which explicitly models the cross-modal interactions between the image and text representations obtained using Contrastive Language-Image Pre-training (CLIP) encoders via a feature interaction matrix (FIM). A simple classifier based on the FIM representation is able to achieve state-of-the-art performance on the Hateful Memes Challenge (HMC) dataset with an AUROC of 85.8, which even surpasses the human performance of 82.65. Experiments on other meme datasets such as Propaganda Memes and TamilMemes also demonstrate the generalizability of the proposed approach. Finally, we analyze the interpretability of the FIM representation and show that cross-modal interactions can indeed facilitate the learning of meaningful concepts. The code for this work is available at https://github.com/gokulkarthik/hateclipper.

Self-omics: A Self-supervised Learning Framework for Multi-omics Cancer Data

We have gained access to vast amounts of multi-omics data thanks to Next Generation Sequencing. However, it is challenging to analyse this data due to its high dimensionality and much of it not being annotated. Lack of annotated data is a significant problem in machine learning, and Self-Supervised Learning (SSL) methods are typically used to deal with limited labelled data. However, there is a lack of studies that use SSL methods to exploit inter-omics relationships on unlabelled multi-omics data. In this work, we develop a novel and efficient pre-training paradigm that consists of various SSL components, including but not limited to contrastive alignment, data recovery from corrupted samples, and using one type of omics data to recover other omic types. Our pre-training paradigm improves performance on downstream tasks with limited labelled data. We show that our approach outperforms the state-of-the-art method in cancer type classification on the TCGA pan-cancer dataset in semi-supervised setting. Moreover, we show that the encoders that are pre-trained using our approach can be used as powerful feature extractors even without fine-tuning. Our ablation study shows that the method is not overly dependent on any pretext task component. The network architectures in our approach are designed to handle missing omic types and multiple datasets for pre-training and downstream training. Our pre-training paradigm can be extended to perform zero-shot classification of rare cancers.

Learning an Ensemble of Deep Fingerprint Representations

Deep neural networks (DNNs) have shown incredible promise in learning fixed-length representations from fingerprints. Since the representation learning is often focused on capturing specific prior knowledge (e.g., minutiae), there is no universal representation that comprehensively encapsulates all the discriminatory information available in a fingerprint. While learning an ensemble of representations can mitigate this problem, two critical challenges need to be addressed: (i) How to extract multiple diverse representations from the same fingerprint image? and (ii) How to optimally exploit these representations during the matching process? In this work, we train multiple instances of DeepPrint (a state-of-the-art DNN-based fingerprint encoder) on different transformations of the input image to generate an ensemble of fingerprint embeddings. We also propose a feature fusion technique that distills these multiple representations into a single embedding, which faithfully captures the diversity present in the ensemble without increasing the computational complexity. The proposed approach has been comprehensively evaluated on five databases containing rolled, plain, and latent fingerprints (NIST SD4, NIST SD14, NIST SD27, NIST SD302, and FVC2004 DB2A) and statistically significant improvements in accuracy have been consistently demonstrated across a range of verification as well as closed- and open-set identification settings. The proposed approach serves as a wrapper capable of improving the accuracy of any DNN-based recognition system.