Abstract

The importance of Scene Text Recognition (STR) in today’s increasingly digital world cannot be overstated. Given the significance of STR, data-intensive deep learning approaches that auto-learn feature mappings have primarily driven the development of STR solutions. Several benchmark datasets and substantial work on deep learning models are available for Latin languages to meet this need. On more complex, syntactically and semantically, Indian languages spoken and read by 1.3 billion people, there is less work and datasets available. This paper aims to address the Indian space’s lack of a comprehensive dataset by proposing the largest and most comprehensive real dataset - IndicSTR12 - and benchmarking STR performance on 12 major Indian languages (Assamese, Bengali, Odia, Marathi, Hindi, Kannada, Urdu, Telugu, Malayalam, Tamil, Gujarati, and Punjabi). A few works have addressed the same …

Abstract

Individual differences are known to modulate brain responses to music. Recent neuroscience research suggests that each individual has unique and fundamentally stable functional brain connections irrespective of task. Our study aims to identify individual differences such as gender and musical expertise from brain responses to music. Thirty-six participants’ functional Magnetic Resonance Imaging (fMRI) responses were measured while listening to three 8-min-long musical pieces representing different musical styles. They were analyzed using various temporal and spectral functional connectivity (FC) measures. These FC measures were compared to identify the one that captured the most variation associated with gender and musical expertise. Subsequently, the measures were used to classify distinct population groupings using a binary Support Vector Machine (SVM). Our results revealed that Coherence, a spectral-domain FC measure, captured the maximum variation for musical stimuli. However, in classifying individuals based on gender and musical expertise, a composite measure outperformed any single measure and had abovechance accuracy. This suggests that each FC measure captures different aspects of the relationship between brain regions and is influenced by the level of analysis (group or individual) and the task, such as similarity analysis or classification. Further ramifications of the findings are discussed.

Abstract

Localizing document objects such as tables, figures, and equations is a primary step for extracting information from document images. We propose a novel end-to-end trainable deep network, termed Document Object Localization Network (dolnet), for detecting various objects present in the document images. The proposed network is a multi-stage extension of Mask r-cnn with a dual backbone having deformable convolution for detecting document objects with high detection accuracy at a higher IoU threshold. We also empirically evaluate the proposed dolnet on the publicly available benchmark datasets. The proposed DOLNet achieves state-of-the-art performance for most of the bench-mark datasets under various existing experimental environments. Our solution has three important properties: (i) a single trained model dolnet that performs well across all the popular benchmark datasets, (ii) reports excellent performances across multiple, including with higher IoU thresholds, and (iii) consistently demonstrate the superior quantitative performance by following the same protocol of the recent works for each of the benchmarks.

Abstract

Sexism, a form of oppression based on one's sex, manifests itself in numerous ways and causes enormous suffering. In view of the growing number of experiences of sexism reported online, automatically classifying these recollections can aid in the battle against sexism by allowing gender studies researchers and government officials involved in policymaking to conduct more effective analyses. This paper investigates the 23-class fine- grained, multi-label classification of accounts (reports) of sexism.

Abstract

This work proposes the utilization of a self-supervised pretrained network for developing a Language Identification (LID) system catering to low-resource Indian languages. The framework employed is Wav2vec2.0-XLSR-53, pre-trained on 53k hours of unlabeled speech data. The unsupervised training of the model enables it to learn the acoustic patterns specific to a language. Given that languages share phonetic space, multi-lingual pre-training is instrumental in learning crosslingual information and building systems that cater to low-resource languages. The results showcase a relative improvement of 33.2% over the DNN-A (DNN with attention) model and 19.04% over Dense Resnets for the Language Identification task on the IIITH-ILSC database using the proposed features.

Abstract

This paper focuses on the Multi-Object Navigation (MultiON) task, which requires a robot to localize an instance (each) of multiple object classes. This is a fundamental task for an assistive robot in a home or a factory. Existing methods for this task have viewed this as a direct extension of Object Navigation (ON), the task of localising a single instance of an object class, and are pre-sequenced, i.e., the sequence in which the object classes are to be explored is provided in advance. This is a strong limitation in practical applications characterized by dynamic changes. On the other hand, our deep reinforcement learning framework for sequence-agnostic MultiON is based on an actor-critic architecture and a reward specification. It exploits past experiences and seeks to reward progress towards individual as well as multiple target object classes. We use photo-realistic scenes from Gibson benchmark dataset in the AI Habitat 3D simulation environment to experimentally show that our method performs better than a pre-sequenced approach and a state of the art ON method extended to MultiON.

Abstract

We present a new class of private information retrieval (PIR) schemes that keep the identity of the file requested private in the presence of at most t colluding servers, based on the recent framework developed for such t-PIR schemes using star products of transitive codes. These t-PIR schemes employ the class of Berman codes as the storage-retrieval code pairs. Berman codes, which are binary linear codes of length nm for any n ≥ 2 and m ≥ 1 being positive integers, were recently shown to achieve the capacity of the binary erasure channel. We provide a complete characterization of the star products of the Berman code pairs, enabling us to calculate the PIR rate of the star product-based schemes that employ these codes. The schemes we present have flexibility in the number of servers, the PIR rate, the storage rate, and the collusion parameter t, owing to numerous codes available in the class of Berman codes. Due to space restrictions, the full version of this paper containing proofs of claims, additional examples, and results,

Abstract

Recent developments in mmWave radar technologies have enabled the truly non-contact heart-rate (HR) and breath-rate (BR) measurement approaches, which provides a great ease in patient monitoring. Additionally, these technologies also provide opportunities to simultaneously detect HR and BR of multiple patients, which has become increasingly important for efficient mass monitoring scenarios. In this work, a frequency modulated continuous wave (FMCW) mmWave radar based truly non-contact multiple patient HR and BR monitoring system has been presented. Furthermore, a novel approach is also proposed, which combines multiple processing methods using a least squares solution to improve measurement accuracy, generalization, and handle measurement error. The proposed system has been developed using Texas Instruments’ FMCW radar and experimental results with multiple subjects are also presented, which show >97% and >93% accuracy in the measured BR and HR values, respectively.

Abstract

Recovering full 3D shapes from partial observations is a challenging task that has been extensively addressed in the computer vision community. Many deep learning methods tackle this problem by training 3D shape generation networks to learn a prior over the full 3D shapes. In this training regime, the methods expect the inputs to be in a fixed canonical form, without which they fail to learn a valid prior over the 3D shapes. We propose SCARP, a model that performs Shape C ompletion in ARbitrary Poses. Given a partial pointcloud of an object, SCARP learns a disentangled feature representation of pose and shape by relying on rotationally equivariant pose features and geometric shape features trained using a multi-tasking objective. Unlike existing methods that depend on an external canonicalization method, SCARP performs canonicalization, pose estimation, and shape completion in a single network, improving the performance by 45% over the existing baselines. In this work, we use SCARP for improving grasp proposals on tabletop objects. By completing partial tabletop objects directly in their observed poses, SCARP enables a SOTA grasp proposal network improve their proposals by 71.2% on partial shapes. Project page: https://bipashasen.github.io/scarp

Abstract

The paper presents a novel nW range gate-leakage-based Sub-Bandgap Voltage Reference (sub-BGR) for low-power and high-temperature range IoT applications. It generates a reference voltage of 336mV without incorporating any resistors and operating for a high-temperature range of −40°C to 150°C and a supply range of 0.7V-4V. In the above temperature and supply ranges, the proposed circuit's power consumption only goes up by 30x and 1.025x times, respectively. Designed in a 65nm CMOS process, the proposed architecture achieves an accuracy of 94ppm/°C. It achieves a line sensitivity of 0.0066%/V for a supply range of 0.7V to 4V and a PSRR of 89dB at DC and 1V supply. The proposed circuit shows a±3σ -inaccuracy of 4.295% without additional trimming circuits. It occupies only 0.0851mm2 area while consuming only 2.3nW at 27°C and 21.74nW at 150°C for a 0.7V supply.