Abstract:Sign languages are the primary means of communication for many hard-of-hearing people worldwide. Recently, to bridge the communication gap between the hard-of-hearing community and the rest of the population, several sign language translation datasets have been proposed to enable the development of statistical sign language translation systems. However, there is a dearth of sign language resources for the Indian sign language. This resource paper introduces ISLTranslate, a translation dataset for continuous Indian Sign Language (ISL) consisting of 31k ISL-English sentence/phrase pairs. To the best of our knowledge, it is the largest translation dataset for continuous Indian Sign Language. We provide a detailed analysis of the dataset. To validate the performance of existing end-to-end Sign language to spoken language translation systems, we benchmark the created dataset with a transformer-based model for ISL translation.
Abstract:Progress in GANs has enabled the generation of high-resolution photorealistic images of astonishing quality. StyleGANs allow for compelling attribute modification on such images via mathematical operations on the latent style vectors in the W/W+ space that effectively modulate the rich hierarchical representations of the generator. Such operations have recently been generalized beyond mere attribute swapping in the original StyleGAN paper to include interpolations. In spite of many significant improvements in StyleGANs, they are still seen to generate unnatural images. The quality of the generated images is predicated on two assumptions; (a) The richness of the hierarchical representations learnt by the generator, and, (b) The linearity and smoothness of the style spaces. In this work, we propose a Hierarchical Semantic Regularizer (HSR) which aligns the hierarchical representations learnt by the generator to corresponding powerful features learnt by pretrained networks on large amounts of data. HSR is shown to not only improve generator representations but also the linearity and smoothness of the latent style spaces, leading to the generation of more natural-looking style-edited images. To demonstrate improved linearity, we propose a novel metric - Attribute Linearity Score (ALS). A significant reduction in the generation of unnatural images is corroborated by improvement in the Perceptual Path Length (PPL) metric by 16.19% averaged across different standard datasets while simultaneously improving the linearity of attribute-change in the attribute editing tasks.
Abstract:We present a motion segmentation guided convolutional neural network (CNN) approach for high dynamic range (HDR) image deghosting. First, we segment the moving regions in the input sequence using a CNN. Then, we merge static and moving regions separately with different fusion networks and combine fused features to generate the final ghost-free HDR image. Our motion segmentation guided HDR fusion approach offers significant advantages over existing HDR deghosting methods. First, by segmenting the input sequence into static and moving regions, our proposed approach learns effective fusion rules for various challenging saturation and motion types. Second, we introduce a novel memory network that accumulates the necessary features required to generate plausible details in the saturated regions. The proposed method outperforms nine existing state-of-the-art methods on two publicly available datasets and generates visually pleasing ghost-free HDR results. We also present a large-scale motion segmentation dataset of 3683 varying exposure images to benefit the research community.
Abstract:In this work, we introduce LEAD, an approach to discover landmarks from an unannotated collection of category-specific images. Existing works in self-supervised landmark detection are based on learning dense (pixel-level) feature representations from an image, which are further used to learn landmarks in a semi-supervised manner. While there have been advances in self-supervised learning of image features for instance-level tasks like classification, these methods do not ensure dense equivariant representations. The property of equivariance is of interest for dense prediction tasks like landmark estimation. In this work, we introduce an approach to enhance the learning of dense equivariant representations in a self-supervised fashion. We follow a two-stage training approach: first, we train a network using the BYOL objective which operates at an instance level. The correspondences obtained through this network are further used to train a dense and compact representation of the image using a lightweight network. We show that having such a prior in the feature extractor helps in landmark detection, even under drastically limited number of annotations while also improving generalization across scale variations.
Abstract:Recent algorithms for image manipulation detection almost exclusively use deep network models. These approaches require either dense pixelwise groundtruth masks, camera ids, or image metadata to train the networks. On one hand, constructing a training set to represent the countless tampering possibilities is impractical. On the other hand, social media platforms or commercial applications are often constrained to remove camera ids as well as metadata from images. A self-supervised algorithm for training manipulation detection models without dense groundtruth or camera/image metadata would be extremely useful for many forensics applications. In this paper, we propose self-supervised approach for training splicing detection/localization models from frequency transforms of images. To identify the spliced regions, our deep network learns a representation to capture an image specific signature by enforcing (image) self consistency . We experimentally demonstrate that our proposed model can yield similar or better performances of multiple existing methods on standard datasets without relying on labels or metadata.
Abstract:We propose a novel recurrent network-based HDR deghosting method for fusing arbitrary length dynamic sequences. The proposed method uses convolutional and recurrent architectures to generate visually pleasing, ghosting-free HDR images. We introduce a new recurrent cell architecture, namely Self-Gated Memory (SGM) cell, that outperforms the standard LSTM cell while containing fewer parameters and having faster running times. In the SGM cell, the information flow through a gate is controlled by multiplying the gate's output by a function of itself. Additionally, we use two SGM cells in a bidirectional setting to improve output quality. The proposed approach achieves state-of-the-art performance compared to existing HDR deghosting methods quantitatively across three publicly available datasets while simultaneously achieving scalability to fuse variable-length input sequence without necessitating re-training. Through extensive ablations, we demonstrate the importance of individual components in our proposed approach. The code is available at https://val.cds.iisc.ac.in/HDR/HDRRNN/index.html.
Abstract:In rural regions of several developing countries, access to quality healthcare, medical infrastructure, and professional diagnosis is largely unavailable. Many of these regions are gradually gaining access to internet infrastructure, although not with a strong enough connection to allow for sustained communication with a medical practitioner. Several deaths resulting from this lack of medical access, absence of patient's previous health records, and the unavailability of information in indigenous languages can be easily prevented. In this paper, we describe an approach leveraging the phenomenal progress in Machine Learning and NLP (Natural Language Processing) techniques to design a model that is low-resource, multilingual, and a preliminary first-point-of-contact medical assistant. Our contribution includes defining the NLP pipeline required for named-entity-recognition, language-agnostic sentence embedding, natural language translation, information retrieval, question answering, and generative pre-training for final query processing. We obtain promising results for this pipeline and preliminary results for EHR (Electronic Health Record) analysis with text summarization for medical practitioners to peruse for their diagnosis. Through this NLP pipeline, we aim to provide preliminary medical information to the user and do not claim to supplant diagnosis from qualified medical practitioners. Using the input from subject matter experts, we have compiled a large corpus to pre-train and fine-tune our BioBERT based NLP model for the specific tasks. We expect recent advances in NLP architectures, several of which are efficient and privacy-preserving models, to further the impact of our solution and improve on individual task performance.