Abstract

Humans involuntarily tend to infer parts of the conversation from lip movements when the speech is absent or corrupted by external noise. In this work, we explore the task of lip to speech synthesis, i.e., learning to generate natural speech given only the lip movements of a speaker. Acknowledging the importance of contextual and speakerspecific cues for accurate lip-reading, we take a different path from existing works. We focus on learning accurate lip sequences to speech mappings for individual speakers in unconstrained, large vocabulary settings. To this end, we collect and release a large-scale benchmark dataset, the first of its kind, specifically to train and evaluate the singlespeaker lip to speech task in natural settings. We propose a novel approach with key design choices to achieve accurate, natural lip to speech synthesis in such unconstrained scenarios for the first time. Extensive evaluation using quantitative, qualitative metrics and human evaluation shows that our method is four times more intelligible than previous works in this space.

Abstract

No Results Found

Abstract

We present sentence aligned parallel corpora across 10 Indian Languages - Hindi, Telugu, Tamil, Malayalam, Gujarati, Urdu, Bengali, Oriya, Marathi, Punjabi, and English - many of which are categorized as low resource. The corpora are compiled from online sources which have content shared across languages. The corpora presented significantly extends present resources that are either not large enough or are restricted to a specific domain (such as health). We also provide a separate test corpus compiled from an independent online source that can be independently used for validating the performance in 10 Indian languages. Alongside, we report on the methods of constructing such corpora using tools enabled by recent advances in machine translation and cross-lingual retrieval using deep neural network based methods.

Abstract

India is a country where several tens of languages are spoken by over a billion strong population. Text-to-speech systems for such languages will thus be extremely beneficial for wide-spread content creation and accessibility. Despite this, the current TTS systems for even the most popular Indian languages fall short of the contemporary state-of-the-art systems for English, Chinese, etc. We believe that one of the major reasons for this is the lack of large, publicly available text-to-speech corpora in these languages that are suitable for training neural text-to-speech systems. To mitigate this, we release a 24 hour text-to-speech corpus for 3 major Indian languages namely Hindi, Malayalam and Bengali. In this work, we also train a state-of-the-art TTS system for each of these languages and report their performances. The collected corpus, code, and trained models are made publicly available.

Abstract

The computationally expensive nature of ab initio molecular dynamics simulations severely limits its ability to simulate large system sizes and long time scales, both of which are necessary to imitate experimental conditions. In this work, we explore an approach to make use of the data obtained using the quantum mechanical density functional theory (DFT) on small systems and use deep learning to subsequently simulate large systems by taking liquid argon as a test case. A suitable vector representation was chosen to represent the surrounding environment of each Ar atom, and a -NetFF machine learning model where, the neural network was trained to predict the di↵erence in resultant forces obtained by DFT and classical force fields was introduced. Molecular dynamics simulations were then performed using forces from the neural network for various system sizes and time scales depending on the properties we calculated. A comparison of properties obtained from the classical force field and the neural network model was presented alongside available experimental data to validate the proposed method.

Abstract

Abstract— Perceiving text is crucial to understand semantics of outdoor scenes and hence is a critical requirement to build intelligent systems for driver assistance and self-driving. Most of the existing datasets for text detection and recognition comprise still images and are mostly compiled keeping text in mind. This paper introduces a new ”RoadText-1K” dataset for text in driving videos. The dataset is 20 times larger than the existing largest dataset for text in videos. Our dataset comprises 1000 video clips of driving without any bias towards text and with annotations for text bounding boxes and transcriptions in every frame. State of the art methods for text detection, recognition and tracking are evaluated on the new dataset and the results signify the challenges in unconstrained driving videos compared to existing datasets. This suggests that RoadText-1K is suited for research and development of reading systems, robust enough to be incorporated into more complex downstream tasks like driver assistance and self-driving. The dataset can be found at http://cvit.iiit.ac.in/research/ projects/cvit-projects/roadtext-1k

Abstract

We present a new dataset for Visual Question Answering on document images called DocVQA. The dataset consists of 50,000 questions defined on 12,000+ document images. We provide detailed analysis of the dataset in comparison with similar datasets for VQA and reading comprehension. We report several baseline results by adopting existing VQA and reading comprehension models. Although the existing models perform reasonably well on certain types of questions, there is large performance gap compared to human performance (94.36% accuracy). The models need to improve specifically on questions where understanding structure of the document in crucial.

Abstract

Recognition and retrieval of textual content from the large document collections have been a powerful use case for the document image analysis community. Often the word is the basic unit for recognition as well as retrieval. Systems that rely only on the text recognisers (ocr) output are not robust enough in many situations, especially when the word recognition rates are poor, as in the case of historic documents or digital libraries. An alternative has been word spotting based methods that retrieve/match words based on a holistic representation of the word. In this paper, we fuse the noisy output of text recogniser with a deep embeddings representation derived out of the entire word. We use average and max fusion for improving the ranked results in the case of retrieval. We validate our methods on a collection of Hindi documents. We improve word recognition rate by 1.4% and retrieval by 11.13% in the mAP.

Abstract

This work explores a hybrid approach to segmentation as an alternative to a purely data-driven approach. We introduce an end-to-end U-Net based network called DU-Net, which uses additional frequency preserving features, namely the Scattering Coefficients (SC), for medical image segmentation. SC are translation invariant and Lipschitz continuous to deformations which help DU-Net outperform other conventional CNN counterparts on four datasets and two segmentation tasks: Optic Disc and Optic Cup in color fundus images and fetal Head in ultrasound images. The proposed method shows remarkable improvement over the basic U-Net with performance competitive to state-of-the-art methods. The results indicate that it is possible to use a lighter network trained with fewer images (without any augmentation) to attain good segmentation results.

Abstract

Autonomous camera systems are often subjected to an optimization/filtering operation to smoothen and stabilize the rough trajectory estimates. Most common filtering techniques do reduce the irregularities in data; however, they fail to mimic the behavior of a human cameraman. Global filtering methods modeling human camera operators have been successful; however, they are limited to offline settings. In this paper, we propose two online filtering methods called Cinefilters, which produce smooth camera trajectories that are motivated by cinematographic principles. The first filter (CineConvex) uses a sliding window-based convex optimization formulation, and the second (CineCNN) is a CNN based encoder-decoder model. We evaluate the proposed filters in two different settings, namely a basketball dataset and a stage performance dataset. Our models outperform previous methods and baselines on both quantitative and qualitative metrics. The CineConvex and CineCNN filters operate at about 250fps and 1000fps, respectively, with a minor latency (half a second), making them apt for a variety of real-time applications.