Bhargava Urala Kota

Srikar Appalaraju, Bhavan Jasani, Bhargava Urala Kota et al.

We present DocFormer -- a multi-modal transformer based architecture for the task of Visual Document Understanding (VDU). VDU is a challenging problem which aims to understand documents in their varied formats (forms, receipts etc.) and layouts. In addition, DocFormer is pre-trained in an unsupervised fashion using carefully designed tasks which encourage multi-modal interaction. DocFormer uses text, vision and spatial features and combines them using a novel multi-modal self-attention layer. DocFormer also shares learned spatial embeddings across modalities which makes it easy for the model to correlate text to visual tokens and vice versa. DocFormer is evaluated on 4 different datasets each with strong baselines. DocFormer achieves state-of-the-art results on all of them, sometimes beating models 4x its size (in no. of parameters).

Muhammet Bastan, Hao-Yu Wu, Tian Cao et al.

We present an open-set logo detection (OSLD) system, which can detect (localize and recognize) any number of unseen logo classes without re-training; it only requires a small set of canonical logo images for each logo class. We achieve this using a two-stage approach: (1) Generic logo detection to detect candidate logo regions in an image. (2) Logo matching for matching the detected logo regions to a set of canonical logo images to recognize them. We constructed an open-set logo detection dataset with 12.1k logo classes and released it for research purposes.We demonstrate the effectiveness of OSLD on our dataset and on the standard Flickr-32 logo dataset, outperforming the state-of-the-art open-set and closed-set logo detection methods by a large margin. OSLD is scalable to millions of logo classes.

Devansh Arpit, Yingbo Zhou, Bhargava U. Kota et al.

While the authors of Batch Normalization (BN) identify and address an important problem involved in training deep networks-- Internal Covariate Shift-- the current solution has certain drawbacks. Specifically, BN depends on batch statistics for layerwise input normalization during training which makes the estimates of mean and standard deviation of input (distribution) to hidden layers inaccurate for validation due to shifting parameter values (especially during initial training epochs). Also, BN cannot be used with batch-size 1 during training. We address these drawbacks by proposing a non-adaptive normalization technique for removing internal covariate shift, that we call Normalization Propagation. Our approach does not depend on batch statistics, but rather uses a data-independent parametric estimate of mean and standard-deviation in every layer thus being computationally faster compared with BN. We exploit the observation that the pre-activation before Rectified Linear Units follow Gaussian distribution in deep networks, and that once the first and second order statistics of any given dataset are normalized, we can forward propagate this normalization without the need for recalculating the approximate statistics for hidden layers.

Rohit Kumar Pandey, Yingbo Zhou, Bhargava Urala Kota et al.

In this paper we present a framework for secure identification using deep neural networks, and apply it to the task of template protection for face authentication. We use deep convolutional neural networks (CNNs) to learn a mapping from face images to maximum entropy binary (MEB) codes. The mapping is robust enough to tackle the problem of exact matching, yielding the same code for new samples of a user as the code assigned during training. These codes are then hashed using any hash function that follows the random oracle model (like SHA-512) to generate protected face templates (similar to text based password protection). The algorithm makes no unrealistic assumptions and offers high template security, cancelability, and state-of-the-art matching performance. The efficacy of the approach is shown on CMU-PIE, Extended Yale B, and Multi-PIE face databases. We achieve high (~95%) genuine accept rates (GAR) at zero false accept rate (FAR) with up to 1024 bits of template security.