Brejesh Lall

Tayeba Qazi, Ayush Maheshwari, Prerana Mukherjee et al.

Thermal imaging offers a powerful alternative to visible-spectrum vision under challenging conditions such as low illumination and adverse weather, yet foundational vision-language models like CLIP fail to align thermal images with textual descriptions due to a fundamental thermal perception gap. We identify three major challenges: the lack of captioned thermal datasets, the inability of standard LLMs to reason about thermal phenomena, and a key representational challenge in thermal imaging where global scene context and object-level heat signatures conflict when learned together in a single embedding space. To address these, we introduce IR-Cap, the first physics-aware thermal captioning pipeline and dataset providing complementary global and fine-grained thermal descriptions across three public benchmarks, and T-CLIP, a decoupled dual-LoRA framework that independently adapts CLIP for scene-level and object-level thermal understanding. T-CLIP achieves consistent improvements over all baselines across three thermal benchmarks in cross-modal retrieval, and we provide an exploratory demonstration of its applicability to text-conditioned thermal image generation.

Darakshan Rashid, Raza Imam, Dwarikanath Mahapatra et al.

Deep neural networks for chest X-ray classification achieve strong average performance, yet often underperform for specific demographic subgroups, raising critical concerns about clinical safety and equity. Existing debiasing methods frequently yield inconsistent improvements across datasets or attain fairness by degrading overall diagnostic utility, treating fairness as a post hoc constraint rather than a property of the learned representation. In this work, we propose Stride-Net (Sensitive Attribute Resilient Learning via Disentanglement and Learnable Masking with Embedding Alignment), a fairness-aware framework that learns disease-discriminative yet demographically invariant representations for chest X-ray analysis. Stride-Net operates at the patch level, using a learnable stride-based mask to select label-aligned image regions while suppressing sensitive attribute information through adversarial confusion loss. To anchor representations in clinical semantics and discourage shortcut learning, we further enforce semantic alignment between image features and BioBERT-based disease label embeddings via Group Optimal Transport. We evaluate Stride-Net on the MIMIC-CXR and CheXpert benchmarks across race and intersectional race-gender subgroups. Across architectures including ResNet and Vision Transformers, Stride-Net consistently improves fairness metrics while matching or exceeding baseline accuracy, achieving a more favorable accuracy-fairness trade-off than prior debiasing approaches. Our code is available at https://github.com/Daraksh/Fairness_StrideNet.

Jyotirmoy Nath, Neeraj Kumar, Brejesh Lall

Automatic prompt optimization (APO) has driven significant gains in LLM-based agentic workflows. However, existing methods treat each task's prompt as a monolithic, instance-blind string optimized through global edits, producing brittle updates and preventing the reuse of learned sub-behaviors. We propose Prompt Codebooks (PCO), a novel compositional prompt optimization framework that recasts APO as discrete learning over a finite vocabulary of natural-language instincts - atomic, reusable instruction units. PCO organizes prompt-construction knowledge in a discrete codebook and routes each input to a small subset of entries via an LLM-based encoder; a generator composes them into a prompt for the frozen target model; a critic emits a structured verdict that decomposes by attribution into per-variable textual gradients, jointly training the encoder, generator, and codebook under a language-valued min-max objective. The resulting routing is per-instance: different inputs in the same task receive different instinct compositions, a regime structurally inexpressible under instance-blind methods. Across six benchmarks on Qwen3-8B and LLaMA-3.1-8B, PCO improves over zero-shot by up to +30.36 points, surpasses the strongest prior baseline (GEPA) by +3.34 on HotpotQA and +1.11 in aggregate, and reduces deployed prompt length by up to 14.1x versus MIPROv2 and 3.0x versus GEPA using only K=16 instincts.

Avinash Upadhyay, Manoj sharma, Prerana Mukherjee et al.

Synthetic infrared (IR) scene and target generation is an important computer vision problem as it allows the generation of realistic IR images and targets for training and testing of various applications, such as remote sensing, surveillance, and target recognition. It also helps reduce the cost and risk associated with collecting real-world IR data. This survey paper aims to provide a comprehensive overview of the conventional mathematical modelling-based methods and deep learning-based methods used for generating synthetic IR scenes and targets. The paper discusses the importance of synthetic IR scene and target generation and briefly covers the mathematics of blackbody and grey body radiations, as well as IR image-capturing methods. The potential use cases of synthetic IR scenes and target generation are also described, highlighting the significance of these techniques in various fields. Additionally, the paper explores possible new ways of developing new techniques to enhance the efficiency and effectiveness of synthetic IR scenes and target generation while highlighting the need for further research to advance this field.

Gousia Habib, Tausifa Jan Saleem, Brejesh Lall

In Natural Language Processing (NLP), Transformers have already revolutionized the field by utilizing an attention-based encoder-decoder model. Recently, some pioneering works have employed Transformer-like architectures in Computer Vision (CV) and they have reported outstanding performance of these architectures in tasks such as image classification, object detection, and semantic segmentation. Vision Transformers (ViTs) have demonstrated impressive performance improvements over Convolutional Neural Networks (CNNs) due to their competitive modelling capabilities. However, these architectures demand massive computational resources which makes these models difficult to be deployed in the resource-constrained applications. Many solutions have been developed to combat this issue, such as compressive transformers and compression functions such as dilated convolution, min-max pooling, 1D convolution, etc. Model compression has recently attracted considerable research attention as a potential remedy. A number of model compression methods have been proposed in the literature such as weight quantization, weight multiplexing, pruning and Knowledge Distillation (KD). However, techniques like weight quantization, pruning and weight multiplexing typically involve complex pipelines for performing the compression. KD has been found to be a simple and much effective model compression technique that allows a relatively simple model to perform tasks almost as accurately as a complex model. This paper discusses various approaches based upon KD for effective compression of ViT models. The paper elucidates the role played by KD in reducing the computational and memory requirements of these models. The paper also presents the various challenges faced by ViTs that are yet to be resolved.

Prashant Pandey, Mustafa Chasmai, Tanuj Sur et al.

Despite the tremendous progress made by deep learning models in image semantic segmentation, they typically require large annotated examples, and increasing attention is being diverted to problem settings like Few-Shot Learning (FSL) where only a small amount of annotation is needed for generalisation to novel classes. This is especially seen in medical domains where dense pixel-level annotations are expensive to obtain. In this paper, we propose Regularized Prototypical Neural Ordinary Differential Equation (R-PNODE), a method that leverages intrinsic properties of Neural-ODEs, assisted and enhanced by additional cluster and consistency losses to perform Few-Shot Segmentation (FSS) of organs. R-PNODE constrains support and query features from the same classes to lie closer in the representation space thereby improving the performance over the existing Convolutional Neural Network (CNN) based FSS methods. We further demonstrate that while many existing Deep CNN based methods tend to be extremely vulnerable to adversarial attacks, R-PNODE exhibits increased adversarial robustness for a wide array of these attacks. We experiment with three publicly available multi-organ segmentation datasets in both in-domain and cross-domain FSS settings to demonstrate the efficacy of our method. In addition, we perform experiments with seven commonly used adversarial attacks in various settings to demonstrate R-PNODE's robustness. R-PNODE outperforms the baselines for FSS by significant margins and also shows superior performance for a wide array of attacks varying in intensity and design.

Sanhita Pathak, Vinay Kaushik, Brejesh Lall

Image-based virtual try-on aims to fit an in-shop garment onto a clothed person image. Garment warping, which aligns the target garment with the corresponding body parts in the person image, is a crucial step in achieving this goal. Existing methods often use multi-stage frameworks to handle clothes warping, person body synthesis and tryon generation separately or rely on noisy intermediate parser-based labels. We propose a novel single-stage framework that implicitly learns the same without explicit multi-stage learning. Our approach utilizes a novel semantic-contextual fusion attention module for garment-person feature fusion, enabling efficient and realistic cloth warping and body synthesis from target pose keypoints. By introducing a lightweight linear attention framework that attends to garment regions and fuses multiple sampled flow fields, we also address misalignment and artifacts present in previous methods. To achieve simultaneous learning of warped garment and try-on results, we introduce a Warped Cloth Learning Module. Our proposed approach significantly improves the quality and efficiency of virtual try-on methods, providing users with a more reliable and realistic virtual try-on experience.

Prashant Pandey, Aleti Vardhan, Mustafa Chasmai et al.

Few-shot Learning (FSL) methods are being adopted in settings where data is not abundantly available. This is especially seen in medical domains where the annotations are expensive to obtain. Deep Neural Networks have been shown to be vulnerable to adversarial attacks. This is even more severe in the case of FSL due to the lack of a large number of training examples. In this paper, we provide a framework to make few-shot segmentation models adversarially robust in the medical domain where such attacks can severely impact the decisions made by clinicians who use them. We propose a novel robust few-shot segmentation framework, Prototypical Neural Ordinary Differential Equation (PNODE), that provides defense against gradient-based adversarial attacks. We show that our framework is more robust compared to traditional adversarial defense mechanisms such as adversarial training. Adversarial training involves increased training time and shows robustness to limited types of attacks depending on the type of adversarial examples seen during training. Our proposed framework generalises well to common adversarial attacks like FGSM, PGD and SMIA while having the model parameters comparable to the existing few-shot segmentation models. We show the effectiveness of our proposed approach on three publicly available multi-organ segmentation datasets in both in-domain and cross-domain settings by attacking the support and query sets without the need for ad-hoc adversarial training.

Prashant Pandey, Mustafa Chasmai, Monish Natarajan et al.

Increasing attention is being diverted to data-efficient problem settings like Open Vocabulary Semantic Segmentation (OVSS) which deals with segmenting an arbitrary object that may or may not be seen during training. The closest standard problems related to OVSS are Zero-Shot and Few-Shot Segmentation (ZSS, FSS) and their Cross-dataset variants where zero to few annotations are needed to segment novel classes. The existing FSS and ZSS methods utilize fully supervised pixel-labelled seen classes to segment unseen classes. Pixel-level labels are hard to obtain, and using weak supervision in the form of inexpensive image-level labels is often more practical. To this end, we propose a novel unified weakly supervised OVSS pipeline that can perform ZSS, FSS and Cross-dataset segmentation on novel classes without using pixel-level labels for either the base (seen) or the novel (unseen) classes in an inductive setting. We propose Weakly-Supervised Language-Guided Segmentation Network (WLSegNet), a novel language-guided segmentation pipeline that i) learns generalizable context vectors with batch aggregates (mean) to map class prompts to image features using frozen CLIP (a vision-language model) and ii) decouples weak ZSS/FSS into weak semantic segmentation and Zero-Shot segmentation. The learned context vectors avoid overfitting on seen classes during training and transfer better to novel classes during testing. WLSegNet avoids fine-tuning and the use of external datasets during training. The proposed pipeline beats existing methods for weak generalized Zero-Shot and weak Few-Shot semantic segmentation by 39 and 3 mIOU points respectively on PASCAL VOC and weak Few-Shot semantic segmentation by 5 mIOU points on MS COCO. On a harder setting of 2-way 1-shot weak FSS, WLSegNet beats the baselines by 13 and 22 mIOU points on PASCAL VOC and MS COCO, respectively.

Gousia Habib, Damandeep Singh, Ishfaq Ahmad Malik et al.

The groundbreaking performance of transformers in Natural Language Processing (NLP) tasks has led to their replacement of traditional Convolutional Neural Networks (CNNs), owing to the efficiency and accuracy achieved through the self-attention mechanism. This success has inspired researchers to explore the use of transformers in computer vision tasks to attain enhanced long-term semantic awareness. Vision transformers (ViTs) have excelled in various computer vision tasks due to their superior ability to capture long-distance dependencies using the self-attention mechanism. Contemporary ViTs like Data Efficient Transformers (DeiT) can effectively learn both global semantic information and local texture information from images, achieving performance comparable to traditional CNNs. However, their impressive performance comes with a high computational cost due to very large number of parameters, hindering their deployment on devices with limited resources like smartphones, cameras, drones etc. Additionally, ViTs require a large amount of data for training to achieve performance comparable to benchmark CNN models. Therefore, we identified two key challenges in deploying ViTs on smaller form factor devices: the high computational requirements of large models and the need for extensive training data. As a solution to these challenges, we propose compressing large ViT models using Knowledge Distillation (KD), which is implemented data-free to circumvent limitations related to data availability. Additionally, we conducted experiments on object detection within the same environment in addition to classification tasks. Based on our analysis, we found that datafree knowledge distillation is an effective method to overcome both issues, enabling the deployment of ViTs on less resourceconstrained devices.

Neeraj Kumar, Ankur Narang, Brejesh Lall

Large pre-trained models, such as Bert, GPT, and Wav2Vec, have demonstrated great potential for learning representations that are transferable to a wide variety of downstream tasks . It is difficult to obtain a large quantity of supervised data due to the limited availability of resources and time. In light of this, a significant amount of research has been conducted in the area of adopting large pre-trained datasets for diverse downstream tasks via fine tuning, linear probing, or prompt tuning in low resource settings. Normalization techniques are essential for accelerating training and improving the generalization of deep neural networks and have been successfully used in a wide variety of applications. A lot of normalization techniques have been proposed but the success of normalization in low resource downstream NLP and speech tasks is limited. One of the reasons is the inability to capture expressiveness by rescaling parameters of normalization. We propose KullbackLeibler(KL) Regularized normalization (KL-Norm) which make the normalized data well behaved and helps in better generalization as it reduces over-fitting, generalises well on out of domain distributions and removes irrelevant biases and features with negligible increase in model parameters and memory overheads. Detailed experimental evaluation on multiple low resource NLP and speech tasks, demonstrates the superior performance of KL-Norm as compared to other popular normalization and regularization techniques.

Prashant Pandey, Himanshu Kumar, Devineni Sri Venkatraya Chowdary et al.

Evolving data streams induce joint nonstationarity in continual semantic segmentation, where semantic classes, input distributions, and supervision availability change simultaneously over time. This setting reflects practical structured prediction systems, yet remains largely unexplored in prior continual learning work, which typically studies these factors in isolation. We formalize continual segmentation under coupled class, domain, and label shifts and investigate learning in heterogeneous dense prediction environments with limited annotations and abundant unlabeled data. To address instability and overfitting arising from few-shot supervision under distribution drift, we introduce gradient-adaptive stabilization, a parameter-wise regularization mechanism implemented via gradient-scaled stochastic perturbations that promotes a principled stability-plasticity tradeoff. We further leverage unlabeled data through semi-supervised learning and introduce prototype anchored supervision that validates pseudo-labels via joint confidence and prototype consistency. Together, these mechanisms enable learning under joint nonstationarity in continual segmentation. Extensive empirical evaluation across class-incremental, domain-incremental, and few-shot regimes demonstrates consistent improvements over prior methods in heterogeneous structured prediction settings. Our results expose fundamental failure modes of existing continual segmentation approaches and provide insight into learning robust dense predictors in dynamically evolving environments.

Nilanjana Chatterjee, Sidharatha Garg, A V Subramanyam et al.

Text-Based Person Search (TBPS) has seen significant progress with vision-language models (VLMs), yet it remains constrained by limited training data and the fact that VLMs are not inherently pre-trained for pedestrian-centric recognition. Existing TBPS methods therefore rely on dataset-centric fine-tuning to handle distribution shift, resulting in multiple independently trained models for different datasets. While synthetic data can increase the scale needed to fine-tune VLMs, it does not eliminate dataset-specific adaptation. This motivates a fundamental question: can we train a single unified TBPS model across multiple datasets? We show that naive joint training over all datasets remains sub-optimal because current training paradigms do not scale to a large number of unique person identities and are vulnerable to noisy image-text pairs. To address these challenges, we propose Scale-TBPS with two contributions: (i) a noise-aware unified dataset curation strategy that cohesively merges diverse TBPS datasets; and (ii) a scalable discriminative identity learning framework that remains effective under a large number of unique identities. Extensive experiments on CUHK-PEDES, ICFG-PEDES, RSTPReid, IIITD-20K, and UFine6926 demonstrate that a single Scale-TBPS model outperforms dataset-centric optimized models and naive joint training.

Neeraj Kumar, Ankur Narang, Brejesh Lall

In this paper, we present a Diffusion GAN based approach (Prosodic Diff-TTS) to generate the corresponding high-fidelity speech based on the style description and content text as an input to generate speech samples within only 4 denoising steps. It leverages the novel conditional prosodic layer normalization to incorporate the style embeddings into the multi head attention based phoneme encoder and mel spectrogram decoder based generator architecture to generate the speech. The style embedding is generated by fine tuning the pretrained BERT model on auxiliary tasks such as pitch, speaking speed, emotion,gender classifications. We demonstrate the efficacy of our proposed architecture on multi-speaker LibriTTS and PromptSpeech datasets, using multiple quantitative metrics that measure generated accuracy and MOS.

Gousia Habib, Tausifa Jan Saleem, Ishfaq Ahmad Malik et al.

With the rapid development of computer vision, Vision Transformers (ViTs) offer the tantalising prospect of unified information processing across visual and textual domains due to the lack of inherent inductive biases in ViTs. ViTs require enormous datasets for training. We introduce an innovative ensemble-based distillation approach that distils inductive bias from complementary lightweight teacher models to make their applications practical. Prior systems relied solely on convolution-based teaching. However, this method incorporates an ensemble of light teachers with different architectural tendencies, such as convolution and involution, to jointly instruct the student transformer. Because of these unique inductive biases, instructors can accumulate a wide range of knowledge, even from readily identifiable stored datasets, which leads to enhanced student performance. Our proposed framework LIB-KD also involves precomputing and keeping logits in advance, essentially the unnormalized predictions of the model. This optimisation can accelerate the distillation process by eliminating the need for repeated forward passes during knowledge distillation, significantly reducing the computational burden and enhancing efficiency.

Prerana Mukherjee, Vinay Kaushik, Ronak Gupta et al.

Predicting attributes in the landmark free facial images is itself a challenging task which gets further complicated when the face gets occluded due to the usage of masks. Smart access control gates which utilize identity verification or the secure login to personal electronic gadgets may utilize face as a biometric trait. Particularly, the Covid-19 pandemic increasingly validates the essentiality of hygienic and contactless identity verification. In such cases, the usage of masks become more inevitable and performing attribute prediction helps in segregating the target vulnerable groups from community spread or ensuring social distancing for them in a collaborative environment. We create a masked face dataset by efficiently overlaying masks of different shape, size and textures to effectively model variability generated by wearing mask. This paper presents a deep Multi-Task Learning (MTL) approach to jointly estimate various heterogeneous attributes from a single masked facial image. Experimental results on benchmark face attribute UTKFace dataset demonstrate that the proposed approach supersedes in performance to other competing techniques. The source code is available at https://github.com/ritikajha/Attribute-prediction-in-masked-facial-images-with-deep-multitask-learning

Sushant Dave, Arun Kumar Singh, Prathosh A. P. et al.

This paper describes neural network based approaches to the process of the formation and splitting of word-compounding, respectively known as the Sandhi and Vichchhed, in Sanskrit language. Sandhi is an important idea essential to morphological analysis of Sanskrit texts. Sandhi leads to word transformations at word boundaries. The rules of Sandhi formation are well defined but complex, sometimes optional and in some cases, require knowledge about the nature of the words being compounded. Sandhi split or Vichchhed is an even more difficult task given its non uniqueness and context dependence. In this work, we propose the route of formulating the problem as a sequence to sequence prediction task, using modern deep learning techniques. Being the first fully data driven technique, we demonstrate that our model has an accuracy better than the existing methods on multiple standard datasets, despite not using any additional lexical or morphological resources. The code is being made available at https://github.com/IITD-DataScience/Sandhi_Prakarana

Gousia Habib, Tausifa jan Saleem, Sheikh Musa Kaleem et al.

Deep learning techniques have been demonstrated to surpass preceding cutting-edge machine learning techniques in recent years, with computer vision being one of the most prominent examples. However, deep learning models suffer from significant drawbacks when deployed in resource-constrained environments due to their large model size and high complexity. Knowledge Distillation is one of the prominent solutions to overcome this challenge. This review paper examines the current state of research on knowledge distillation, a technique for compressing complex models into smaller and simpler ones. The paper provides an overview of the major principles and techniques associated with knowledge distillation and reviews the applications of knowledge distillation in the domain of computer vision. The review focuses on the benefits of knowledge distillation, as well as the problems that must be overcome to improve its effectiveness.

Sanhita Pathak, Vinay Kaushik, Brejesh Lall

To prevent unauthorized use of text in images, Scene Text Removal (STR) has become a crucial task. It focuses on automatically removing text and replacing it with a natural, text-less background while preserving significant details such as texture, color, and contrast. Despite its importance in privacy protection, STR faces several challenges, including boundary artifacts, inconsistent texture and color, and preserving correct shadows. Most STR approaches estimate a text region mask to train a model, solving for image translation or inpainting to generate a text-free image. Thus, the quality of the generated image depends on the accuracy of the inpainting mask and the generator's capability. In this work, we leverage the superior capabilities of diffusion models in generating high-quality, consistent images to address the STR problem. We introduce a ControlNet diffusion model, treating STR as an inpainting task. To enhance the model's robustness, we develop a mask pretraining pipeline to condition our diffusion model. This involves training a masked autoencoder (MAE) using a combination of box masks and coarse stroke masks, and fine-tuning it using masks derived from our novel segmentation-based mask refinement framework. This framework iteratively refines an initial mask and segments it using the SLIC and Hierarchical Feature Selection (HFS) algorithms to produce an accurate final text mask. This improves mask prediction and utilizes rich textural information in natural scene images to provide accurate inpainting masks. Experiments on the SCUT-EnsText and SCUT-Syn datasets demonstrate that our method significantly outperforms existing state-of-the-art techniques.

Vinti Nayar, Kanica Sachdev, Brejesh Lall

Quality of Experience (QoE) prediction plays a crucial role in optimizing resource management and enhancing user satisfaction across both telecommunication and OTT services. While recent advances predominantly rely on deep learning models, this study introduces a lightweight generative modeling framework that balances computational efficiency, interpretability, and predictive accuracy. By validating the use of Vector Quantization (VQ) as a preprocessing technique, continuous network features are effectively transformed into discrete categorical symbols, enabling integration with a Hidden Markov Model (HMM) for temporal sequence modeling. This VQ-HMM pipeline enhances the model's capacity to capture dynamic QoE patterns while supporting probabilistic inference on new and unseen data. Experimental results on publicly available time-series datasets incorporating both objective indicators and subjective QoE scores demonstrate the viability of this approach in real-time and resource-constrained environments, where inference latency is also critical. The framework offers a scalable alternative to complex deep learning methods, particularly in scenarios with limited computational resources or where latency constraints are critical.

Sadia Hussain, Brejesh Lall

Hyperspectral imaging (HSI) is a powerful earth observation technology that captures and processes information across a wide spectrum of wavelengths. Hyperspectral imaging provides comprehensive and detailed spectral data that is invaluable for a wide range of reconstruction problems. However due to complexity in analysis it often becomes difficult to handle this data. To address the challenge of handling large number of bands in reconstructing high quality HSI, we propose to form groups of bands. In this position paper we propose a method of selecting diverse bands using determinantal point processes in correlated bands. To address the issue of overlapping bands that may arise from grouping, we use spectral angle mapper analysis. This analysis can be fed to any Machine learning model to enable detailed analysis and monitoring with high precision and accuracy.

Kailas Dayanandan, Nikhil Kumar, Anand Sinha et al.

The dual thinking framework considers fast, intuitive, and slower logical processing. The perception of dual thinking in vision requires images where inferences from intuitive and logical processing differ, and the latter is under-explored in current studies. We introduce a novel adversarial dataset to provide evidence for the dual thinking framework in human vision, which also facilitates the study of the qualitative behavior of deep learning models. Our psychophysical studies show the presence of multiple inferences in rapid succession, and analysis of errors shows that the early stopping of visual processing can result in missing relevant information. MLLMs (Multi-modal Large Language Models) and VLMs (Vision Language Models) have made significant progress in correcting errors in intuitive processing in human vision and showed enhanced performance on images requiring logical processing. However, their improvements in logical processing have not kept pace with their advancements in intuitive processing. In contrast, segmentation models exhibit errors similar to those seen in intuitive human processing and lack understanding of sub-structures, as indicated by errors related to sub-components in identified instances. As AI (Artificial Intelligence)-based systems find increasing applications in safety-critical domains like autonomous driving, the integration of logical processing capabilities becomes essential. This not only enhances performance but also addresses the limitations of scaling-based approaches while ensuring robustness and reliability in real-world environments.

Sanhita Pathak, Vinay Kaushik, Brejesh Lall

Virtual try-on, a rapidly evolving field in computer vision, is transforming e-commerce by improving customer experiences through precise garment warping and seamless integration onto the human body. While existing methods such as TPS and flow address the garment warping but overlook the finer contextual details. In this paper, we introduce a novel graph based warping technique which emphasizes the value of context in garment flow. Our graph based warping module generates warped garment as well as a coarse person image, which is utilised by a simple refinement network to give a coarse virtual tryon image. The proposed work exploits latent diffusion model to generate the final tryon, treating garment transfer as an inpainting task. The diffusion model is conditioned with decoupled cross attention based inversion of visual and textual information. We introduce an occlusion aware warping constraint that generates dense warped garment, without any holes and occlusion. Our method, validated on VITON-HD and Dresscode datasets, showcases substantial state-of-the-art qualitative and quantitative results showing considerable improvement in garment warping, texture preservation, and overall realism.

Tausifa Jan Saleem, Ramanjit Ahuja, Surendra Prasad et al.

Lottery ticket hypothesis for deep neural networks emphasizes the importance of initialization used to re-train the sparser networks obtained using the iterative magnitude pruning process. An explanation for why the specific initialization proposed by the lottery ticket hypothesis tends to work better in terms of generalization (and training) performance has been lacking. Moreover, the underlying principles in iterative magnitude pruning, like the pruning of smaller magnitude weights and the role of the iterative process, lack full understanding and explanation. In this work, we attempt to provide insights into these phenomena by empirically studying the volume/geometry and loss landscape characteristics of the solutions obtained at various stages of the iterative magnitude pruning process.

A V Subramanyam, Niranjan Sundararajan, Vibhu Dubey et al.

Text-to-Image (T2I) ReID has attracted a lot of attention in the recent past. CUHK-PEDES, RSTPReid and ICFG-PEDES are the three available benchmarks to evaluate T2I ReID methods. RSTPReid and ICFG-PEDES comprise of identities from MSMT17 but due to limited number of unique persons, the diversity is limited. On the other hand, CUHK-PEDES comprises of 13,003 identities but has relatively shorter text description on average. Further, these datasets are captured in a restricted environment with limited number of cameras. In order to further diversify the identities and provide dense captions, we propose a novel dataset called IIITD-20K. IIITD-20K comprises of 20,000 unique identities captured in the wild and provides a rich dataset for text-to-image ReID. With a minimum of 26 words for a description, each image is densely captioned. We further synthetically generate images and fine-grained captions using Stable-diffusion and BLIP models trained on our dataset. We perform elaborate experiments using state-of-art text-to-image ReID models and vision-language pre-trained models and present a comprehensive analysis of the dataset. Our experiments also reveal that synthetically generated data leads to a substantial performance improvement in both same dataset as well as cross dataset settings. Our dataset is available at https://bit.ly/3pkA3Rj.

Bishshoy Das, Milton Mondal, Brejesh Lall et al.

In standard neural network training, the gradients in the backward pass are determined by the forward pass. As a result, the two stages are coupled. This is how most neural networks are trained currently. However, gradient modification in the backward pass has seldom been studied in the literature. In this paper we explore decoupled training, where we alter the gradients in the backward pass. We propose a simple yet powerful method called PowerGrad Transform, that alters the gradients before the weight update in the backward pass and significantly enhances the predictive performance of the neural network. PowerGrad Transform trains the network to arrive at a better optima at convergence. It is computationally extremely efficient, virtually adding no additional cost to either memory or compute, but results in improved final accuracies on both the training and test sets. PowerGrad Transform is easy to integrate into existing training routines, requiring just a few lines of code. PowerGrad Transform accelerates training and makes it possible for the network to better fit the training data. With decoupled training, PowerGrad Transform improves baseline accuracies for ResNet-50 by 0.73%, for SE-ResNet-50 by 0.66% and by more than 1.0% for the non-normalized ResNet-18 network on the ImageNet classification task.

Muneeb Ahmed, Rajesh Kumar, Qaim Abbas et al.

The letter focuses on Haptic Glove (HG) based control of a Robotic Hand (RH) executing in-hand manipulation of certain objects of interest. The high dimensional motion signals in HG and RH possess intrinsic variability of kinematics resulting in difficulty to establish a direct mapping of the motion signals from HG onto the RH. An estimation mechanism is proposed to quantify the motion signal acquired from the human controller in relation to the intended goal pose of the object being held by the robotic hand. A control algorithm is presented to transform the synthesized intent at the RH and allow relocation of the object to the expected goal pose. The lag in synthesis of the intent in the presence of communication delay leads to a requirement of predicting the estimated intent. We leverage an attention-based convolutional neural network encoder to predict the trajectory of intent for a certain lookahead to compensate for the delays. The proposed methodology is evaluated across objects of different shapes, mass, and materials. We present a comparative performance of the estimation and prediction mechanisms on 5G-driven real-world robotic setup against benchmark methodologies. The test-MSE in prediction of human intent is reported to yield ~ 97.3 -98.7% improvement of accuracy in comparison to LSTM-based benchmark

Vinay Kaushik, Kartik Jindgar, Brejesh Lall

Self-supervised learning of depth has been a highly studied topic of research as it alleviates the requirement of having ground truth annotations for predicting depth. Depth is learnt as an intermediate solution to the task of view synthesis, utilising warped photometric consistency. Although it gives good results when trained using stereo data, the predicted depth is still sensitive to noise, illumination changes and specular reflections. Also, occlusion can be tackled better by learning depth from a single camera. We propose ADAA, utilising depth augmentation as depth supervision for learning accurate and robust depth. We propose a relational self-attention module that learns rich contextual features and further enhances depth results. We also optimize the auto-masking strategy across all losses by enforcing L1 regularisation over mask. Our novel progressive training strategy first learns depth at a lower resolution and then progresses to the original resolution with slight training. We utilise a ResNet18 encoder, learning features for prediction of both depth and pose. We evaluate our predicted depth on the standard KITTI driving dataset and achieve state-of-the-art results for monocular depth estimation whilst having significantly lower number of trainable parameters in our deep learning framework. We also evaluate our model on Make3D dataset showing better generalization than other methods.

Neeraj Kumar, Srishti Goel, Ankur Narang et al.

We consider the challenging problem of audio to animated video generation. We propose a novel method OneShotAu2AV to generate an animated video of arbitrary length using an audio clip and a single unseen image of a person as an input. The proposed method consists of two stages. In the first stage, OneShotAu2AV generates the talking-head video in the human domain given an audio and a person's image. In the second stage, the talking-head video from the human domain is converted to the animated domain. The model architecture of the first stage consists of spatially adaptive normalization based multi-level generator and multiple multilevel discriminators along with multiple adversarial and non-adversarial losses. The second stage leverages attention based normalization driven GAN architecture along with temporal predictor based recycle loss and blink loss coupled with lipsync loss, for unsupervised generation of animated video. In our approach, the input audio clip is not restricted to any specific language, which gives the method multilingual applicability. OneShotAu2AV can generate animated videos that have: (a) lip movements that are in sync with the audio, (b) natural facial expressions such as blinks and eyebrow movements, (c) head movements. Experimental evaluation demonstrates superior performance of OneShotAu2AV as compared to U-GAT-IT and RecycleGan on multiple quantitative metrics including KID(Kernel Inception Distance), Word error rate, blinks/sec

Neeraj Kumar, Srishti Goel, Ankur Narang et al.

Speech-driven facial video generation has been a complex problem due to its multi-modal aspects namely audio and video domain. The audio comprises lots of underlying features such as expression, pitch, loudness, prosody(speaking style) and facial video has lots of variability in terms of head movement, eye blinks, lip synchronization and movements of various facial action units along with temporal smoothness. Synthesizing highly expressive facial videos from the audio input and static image is still a challenging task for generative adversarial networks. In this paper, we propose a multi-modal adaptive normalization(MAN) based architecture to synthesize a talking person video of arbitrary length using as input: an audio signal and a single image of a person. The architecture uses the multi-modal adaptive normalization, keypoint heatmap predictor, optical flow predictor and class activation map[58] based layers to learn movements of expressive facial components and hence generates a highly expressive talking-head video of the given person. The multi-modal adaptive normalization uses the various features of audio and video such as Mel spectrogram, pitch, energy from audio signals and predicted keypoint heatmap/optical flow and a single image to learn the respective affine parameters to generate highly expressive video. Experimental evaluation demonstrates superior performance of the proposed method as compared to Realistic Speech-Driven Facial Animation with GANs(RSDGAN) [53], Speech2Vid [10], and other approaches, on multiple quantitative metrics including: SSIM (structural similarity index), PSNR (peak signal to noise ratio), CPBD (image sharpness), WER(word error rate), blinks/sec and LMD(landmark distance). Further, qualitative evaluation and Online Turing tests demonstrate the efficacy of our approach.

Neeraj Kumar, Srishti Goel, Ankur Narang et al.

The style of the speech varies from person to person and every person exhibits his or her own style of speaking that is determined by the language, geography, culture and other factors. Style is best captured by prosody of a signal. High quality multi-speaker speech synthesis while considering prosody and in a few shot manner is an area of active research with many real-world applications. While multiple efforts have been made in this direction, it remains an interesting and challenging problem. In this paper, we present a novel few shot multi-speaker speech synthesis approach (FSM-SS) that leverages adaptive normalization architecture with a non-autoregressive multi-head attention model. Given an input text and a reference speech sample of an unseen person, FSM-SS can generate speech in that person's style in a few shot manner. Additionally, we demonstrate how the affine parameters of normalization help in capturing the prosodic features such as energy and fundamental frequency in a disentangled fashion and can be used to generate morphed speech output. We demonstrate the efficacy of our proposed architecture on multi-speaker VCTK and LibriTTS datasets, using multiple quantitative metrics that measure generated speech distortion and MoS, along with speaker embedding analysis of the generated speech vs the actual speech samples.

Anushikha Singh, Brejesh Lall, B. K. Panigrahi et al.

Background and Objective: Tuberculosis (TB) is a significant public health issue and a leading cause of death worldwide. Millions of deaths can be averted by early diagnosis and successful treatment of TB patients. Automated diagnosis of TB holds vast potential to assist medical experts in expediting and improving its diagnosis, especially in developing countries like India, where there is a shortage of trained medical experts and radiologists. To date, several deep learning based methods for automated detection of TB from chest radiographs have been proposed. However, the performance of a few of these methods on the Indian chest radiograph data set has been suboptimal, possibly due to different texture of the lungs on chest radiographs of Indian subjects compared to other countries. Thus deep learning for accurate and automated diagnosis of TB on Indian datasets remains an important subject of research. Methods: The proposed work explores the performance of convolutional neural networks (CNNs) for the diagnosis of TB in Indian chest x-ray images. Three different pre-trained neural network models, AlexNet, GoogLenet, and ResNet are used to classify chest x-ray images into healthy or TB infected. The proposed approach does not require any pre-processing technique. Also, other works use pre-trained NNs as a tool for crafting features and then apply standard classification techniques. However, we attempt an end to end NN model based diagnosis of TB from chest x-rays. The proposed visualization tool can also be used by radiologists in the screening of large datasets. Results: The proposed method achieved 93.40% accuracy with 98.60% sensitivity to diagnose TB for the Indian population. Conclusions: The performance of the proposed method is also tested against techniques described in the literature. The proposed method outperforms the state of art on Indian and Shenzhen datasets.

Anushikha Singh, Brejesh Lall, B. K. Panigrahi et al.

A chest radiograph, commonly called chest x-ray (CxR), plays a vital role in the diagnosis of various lung diseases, such as lung cancer, tuberculosis, pneumonia, and many more. Automated segmentation of the lungs is an important step to design a computer-aided diagnostic tool for examination of a CxR. Precise lung segmentation is considered extremely challenging because of variance in the shape of the lung caused by health issues, age, and gender. The proposed work investigates the use of an efficient deep convolutional neural network for accurate segmentation of lungs from CxR. We attempt an end to end DeepLabv3+ network which integrates DeepLab architecture, encoder-decoder, and dilated convolution for semantic lung segmentation with fast training and high accuracy. We experimented with the different pre-trained base networks: Resnet18 and Mobilenetv2, associated with the Deeplabv3+ model for performance analysis. The proposed approach does not require any pre-processing technique on chest x-ray images before being fed to a neural network. Morphological operations were used to remove false positives that occurred during semantic segmentation. We construct a CxR dataset of the Indian population that contain healthy and unhealthy CxRs of clinically confirmed patients of tuberculosis, chronic obstructive pulmonary disease, interstitial lung disease, pleural effusion, and lung cancer. The proposed method is tested on 688 images of our Indian CxR dataset including images with severe abnormal findings to validate its robustness. We also experimented on commonly used benchmark datasets such as Japanese Society of Radiological Technology; Montgomery County, USA; and Shenzhen, China for state-of-the-art comparison. The performance of our method is tested against techniques described in the literature and achieved the highest accuracy for lung segmentation on Indian and public datasets.

Arun Kumar Singh, Sushant Dave, Prathosh A. P. et al.

This paper presents first benchmark corpus of Sanskrit Pratyaya (suffix) and inflectional words (padas) formed due to suffixes along with neural network based approaches to process the formation and splitting of inflectional words. Inflectional words spans the primary and secondary derivative nouns as the scope of current work. Pratyayas are an important dimension of morphological analysis of Sanskrit texts. There have been Sanskrit Computational Linguistics tools for processing and analyzing Sanskrit texts. Unfortunately there has not been any work to standardize & validate these tools specifically for derivative nouns analysis. In this work, we prepared a Sanskrit suffix benchmark corpus called Pratyaya-Kosh to evaluate the performance of tools. We also present our own neural approach for derivative nouns analysis while evaluating the same on most prominent Sanskrit Morphological Analysis tools. This benchmark will be freely dedicated and available to researchers worldwide and we hope it will motivate all to improve morphological analysis in Sanskrit Language.

Ronak Gupta, Prerana Mukherjee, Brejesh Lall et al.

Monument classification can be performed on the basis of their appearance and shape from coarse to fine categories. Although there is much semantic information present in the monuments which is reflected in the eras they were built, its type or purpose, the dynasty which established it, etc. Particularly, Indian subcontinent exhibits a huge deal of variation in terms of architectural styles owing to its rich cultural heritage. In this paper, we propose a framework that utilizes hierarchy to preserve semantic information while performing image classification or image retrieval. We encode the learnt deep embeddings to construct a dictionary of images and then utilize a re-ranking framework on the the retrieved results using DeLF features. The semantic information preserved in these embeddings helps to classify unknown monuments at higher level of granularity in hierarchy. We have curated a large, novel Indian heritage monuments dataset comprising of images of historical, cultural and religious importance with subtypes of eras, dynasties and architectural styles. We demonstrate the performance of the proposed framework in image classification and retrieval tasks and compare it with other competing methods on this dataset.

Kanish Garg, Ajeet kumar Singh, Dorien Herremans et al.

Generating an image from a provided descriptive text is quite a challenging task because of the difficulty in incorporating perceptual information (object shapes, colors, and their interactions) along with providing high relevancy related to the provided text. Current methods first generate an initial low-resolution image, which typically has irregular object shapes, colors, and interaction between objects. This initial image is then improved by conditioning on the text. However, these methods mainly address the problem of using text representation efficiently in the refinement of the initially generated image, while the success of this refinement process depends heavily on the quality of the initially generated image, as pointed out in the DM-GAN paper. Hence, we propose a method to provide good initialized images by incorporating perceptual understanding in the discriminator module. We improve the perceptual information at the first stage itself, which results in significant improvement in the final generated image. In this paper, we have applied our approach to the novel StackGAN architecture. We then show that the perceptual information included in the initial image is improved while modeling image distribution at multiple stages. Finally, we generated realistic multi-colored images conditioned by text. These images have good quality along with containing improved basic perceptual information. More importantly, the proposed method can be integrated into the pipeline of other state-of-the-art text-based-image-generation models to generate initial low-resolution images. We also worked on improving the refinement process in StackGAN by augmenting the third stage of the generator-discriminator pair in the StackGAN architecture. Our experimental analysis and comparison with the state-of-the-art on a large but sparse dataset MS COCO further validate the usefulness of our proposed approach.

Vinay Kaushik, Brejesh Lall

Recent advances in end-to-end unsupervised learning has significantly improved the performance of monocular depth prediction and alleviated the requirement of ground truth depth. Although a plethora of work has been done in enforcing various structural constraints by incorporating multiple losses utilising smoothness, left-right consistency, regularisation and matching surface normals, a few of them take into consideration multi-scale structures present in real world images. Most works utilise a VGG16 or ResNet50 model pre-trained on ImageNet weights for predicting depth. We propose a deep feature fusion method utilising features at multiple scales for learning self-supervised depth from scratch. Our fusion network selects features from both upper and lower levels at every level in the encoder network, thereby creating multiple feature pyramid sub-networks that are fed to the decoder after applying the CoordConv solution. We also propose a refinement module learning higher scale residual depth from a combination of higher level deep features and lower level residual depth using a pixel shuffling framework that super-resolves lower level residual depth. We select the KITTI dataset for evaluation and show that our proposed architecture can produce better or comparable results in depth prediction.

Snigdha Bhagat, S. D. Joshi, Brejesh Lall

It is a challenging task to extract the best of both worlds by combining the spatial characteristics of a visible image and the spectral content of an infrared image. In this work, we propose a spatially constrained adversarial autoencoder that extracts deep features from the infrared and visible images to obtain a more exhaustive and global representation. In this paper, we propose a residual autoencoder architecture, regularised by a residual adversarial network, to generate a more realistic fused image. The residual module serves as primary building for the encoder, decoder and adversarial network, as an add on the symmetric skip connections perform the functionality of embedding the spatial characteristics directly from the initial layers of encoder structure to the decoder part of the network. The spectral information in the infrared image is incorporated by adding the feature maps over several layers in the encoder part of the fusion structure, which makes inference on both the visual and infrared images separately. In order to efficiently optimize the parameters of the network, we propose an adversarial regulariser network which would perform supervised learning on the fused image and the original visual image.

Ronak Gupta, Prashant Anand, Santanu Chaudhury et al.

Fall detection holds immense importance in the field of healthcare, where timely detection allows for instant medical assistance. In this context, we propose a 3D ConvNet architecture which consists of 3D Inception modules for fall detection. The proposed architecture is a custom version of Inflated 3D (I3D) architecture, that takes compressed measurements of video sequence as spatio-temporal input, obtained from compressive sensing framework, rather than video sequence as input, as in the case of I3D convolutional neural network. This is adopted since privacy raises a huge concern for patients being monitored through these RGB cameras. The proposed framework for fall detection is flexible enough with respect to a wide variety of measurement matrices. Ten action classes randomly selected from Kinetics-400 with no fall examples, are employed to train our 3D ConvNet post compressive sensing with different types of sensing matrices on the original video clips. Our results show that 3D ConvNet performance remains unchanged with different sensing matrices. Also, the performance obtained with Kinetics pre-trained 3D ConvNet on compressively sensed fall videos from benchmark datasets is better than the state-of-the-art techniques.

Ajit Jadhav, Prerana Mukherjee, Vinay Kaushik et al.

A lot a research is focused on object detection and it has achieved significant advances with deep learning techniques in recent years. Inspite of the existing research, these algorithms are not usually optimal for dealing with sequences or images captured by drone-based platforms, due to various challenges such as view point change, scales, density of object distribution and occlusion. In this paper, we develop a model for detection of objects in drone images using the VisDrone2019 DET dataset. Using the RetinaNet model as our base, we modify the anchor scales to better handle the detection of dense distribution and small size of the objects. We explicitly model the channel interdependencies by using "Squeeze-and-Excitation" (SE) blocks that adaptively recalibrates channel-wise feature responses. This helps to bring significant improvements in performance at a slight additional computational cost. Using this architecture for object detection, we build a custom DeepSORT network for object detection on the VisDrone2019 MOT dataset by training a custom Deep Association network for the algorithm.

Mohit Goyal, Rajan Goyal, Brejesh Lall

The scope of research in the domain of activation functions remains limited and centered around improving the ease of optimization or generalization quality of neural networks (NNs). However, to develop a deeper understanding of deep learning, it becomes important to look at the non linear component of NNs more carefully. In this paper, we aim to provide a generic form of activation function along with appropriate mathematical grounding so as to allow for insights into the working of NNs in future. We propose "Self-Learnable Activation Functions" (SLAF), which are learned during training and are capable of approximating most of the existing activation functions. SLAF is given as a weighted sum of pre-defined basis elements which can serve for a good approximation of the optimal activation function. The coefficients for these basis elements allow a search in the entire space of continuous functions (consisting of all the conventional activations). We propose various training routines which can be used to achieve performance with SLAF equipped neural networks (SLNNs). We prove that SLNNs can approximate any neural network with lipschitz continuous activations, to any arbitrary error highlighting their capacity and possible equivalence with standard NNs. Also, SLNNs can be completely represented as a collections of finite degree polynomial upto the very last layer obviating several hyper parameters like width and depth. Since the optimization of SLNNs is still a challenge, we show that using SLAF along with standard activations (like ReLU) can provide performance improvements with only a small increase in number of parameters.

Prashant Anand, Ajeet Kumar Singh, Siddharth Srivastava et al.

The recent advances in deep learning are mostly driven by availability of large amount of training data. However, availability of such data is not always possible for specific tasks such as speaker recognition where collection of large amount of data is not possible in practical scenarios. Therefore, in this paper, we propose to identify speakers by learning from only a few training examples. To achieve this, we use a deep neural network with prototypical loss where the input to the network is a spectrogram. For output, we project the class feature vectors into a common embedding space, followed by classification. Further, we show the effectiveness of capsule net in a few shot learning setting. To this end, we utilize an auto-encoder to learn generalized feature embeddings from class-specific embeddings obtained from capsule network. We provide exhaustive experiments on publicly available datasets and competitive baselines, demonstrating the superiority and generalization ability of the proposed few shot learning pipelines.

Divyam Madaan, Radhika Dua, Prerana Mukherjee et al.

Air pollution is the leading environmental health hazard globally due to various sources which include factory emissions, car exhaust and cooking stoves. As a precautionary measure, air pollution forecast serves as the basis for taking effective pollution control measures, and accurate air pollution forecasting has become an important task. In this paper, we forecast fine-grained ambient air quality information for 5 prominent locations in Delhi based on the historical and real-time ambient air quality and meteorological data reported by Central Pollution Control board. We present VayuAnukulani system, a novel end-to-end solution to predict air quality for next 24 hours by estimating the concentration and level of different air pollutants including nitrogen dioxide ($NO_2$), particulate matter ($PM_{2.5}$ and $PM_{10}$) for Delhi. Extensive experiments on data sources obtained in Delhi demonstrate that the proposed adaptive attention based Bidirectional LSTM Network outperforms several baselines for classification and regression models. The accuracy of the proposed adaptive system is $\sim 15 - 20\%$ better than the same offline trained model. We compare the proposed methodology on several competing baselines, and show that the network outperforms conventional methods by $\sim 3 - 5 \%$.

Shiva Azimi, Brejesh lall, Tapan K. Gandhi

Plant Phenomics based on imaging based techniques can be used to monitor the health and the diseases of plants and crops. The use of 3D data for plant phenomics is a recent phenomenon. However, since 3D point cloud contains more information than plant images, in this paper, we compare the performance of different keypoint detectors and local feature descriptors combinations for the plant growth stage and it's growth condition classification based on 3D point clouds of the plants. We have also implemented a modified form of 3D SIFT descriptor, that is invariant to rotation and is computationally less intense than most of the 3D SIFT descriptors reported in the existing literature. The performance is evaluated in terms of the classification accuracy and the results are presented in terms of accuracy tables. We find the ISS-SHOT and the SIFT-SIFT combinations consistently perform better and Fisher Vector (FV) is a better encoder than Vector of Linearly Aggregated (VLAD) for such applications. It can serve as a better modality.

Prerana Mukherjee, Manoj Sharma, Megh Makwana et al.

Synthesizing high quality saliency maps from noisy images is a challenging problem in computer vision and has many practical applications. Samples generated by existing techniques for saliency detection cannot handle the noise perturbations smoothly and fail to delineate the salient objects present in the given scene. In this paper, we present a novel end-to-end coupled Denoising based Saliency Prediction with Generative Adversarial Network (DSAL-GAN) framework to address the problem of salient object detection in noisy images. DSAL-GAN consists of two generative adversarial-networks (GAN) trained end-to-end to perform denoising and saliency prediction altogether in a holistic manner. The first GAN consists of a generator which denoises the noisy input image, and in the discriminator counterpart we check whether the output is a denoised image or ground truth original image. The second GAN predicts the saliency maps from raw pixels of the input denoised image using a data-driven metric based on saliency prediction method with adversarial loss. Cycle consistency loss is also incorporated to further improve salient region prediction. We demonstrate with comprehensive evaluation that the proposed framework outperforms several baseline saliency models on various performance benchmarks.

Siddharth Srivastava, Brejesh Lall

Learning local descriptors is an important problem in computer vision. While there are many techniques for learning local patch descriptors for 2D images, recently efforts have been made for learning local descriptors for 3D points. The recent progress towards solving this problem in 3D leverages the strong feature representation capability of image based convolutional neural networks by utilizing RGB-D or multi-view representations. However, in this paper, we propose to learn 3D local descriptors by directly processing unstructured 3D point clouds without needing any intermediate representation. The method constitutes a deep network for learning permutation invariant representation of 3D points. To learn the local descriptors, we use a multi-margin contrastive loss which discriminates between similar and dissimilar points on a surface while also leveraging the extent of dissimilarity among the negative samples at the time of training. With comprehensive evaluation against strong baselines, we show that the proposed method outperforms state-of-the-art methods for matching points in 3D point clouds. Further, we demonstrate the effectiveness of the proposed method on various applications achieving state-of-the-art results.

Vinay Kaushik, Brejesh Lall

Disparity by Block Matching stereo is usually used in applications with limited computational power in order to get depth estimates. However, the research on simple stereo methods has been lesser than the energy based counterparts which promise a better quality depth map with more potential for future improvements. Semi-global-matching (SGM) methods offer good performance and easy implementation but suffer from the problem of very high memory footprint because it's working on the full disparity space image. On the other hand, Block matching stereo needs much less memory. In this paper, we introduce a novel multi-scale-hierarchical block-matching approach using a pyramidal variant of depth and cost functions which drastically improves the results of standard block matching stereo techniques while preserving the low memory footprint and further reducing the complexity of standard block matching. We tested our new multi block matching scheme on the Middlebury stereo benchmark. For the Middlebury benchmark we get results that are only slightly worse than state of the art SGM implementations.

Vinay Kaushik, Prerana Mukherjee, Brejesh Lall

In this paper, we attempt to advance the research work done in human action recognition to a rather specialized application namely Indian Classical Dance (ICD) classification. The variation in such dance forms in terms of hand and body postures, facial expressions or emotions and head orientation makes pose estimation an extremely challenging task. To circumvent this problem, we construct a pose-oblivious shape signature which is fed to a sequence learning framework. The pose signature representation is done in two-fold process. First, we represent person-pose in first frame of a dance video using symmetric Spatial Transformer Networks (STN) to extract good person object proposals and CNN-based parallel single person pose estimator (SPPE). Next, the pose basis are converted to pose flows by assigning a similarity score between successive poses followed by non-maximal suppression. Instead of feeding a simple chain of joints in the sequence learner which generally hinders the network performance we constitute a feature vector of the normalized distance vectors, flow, angles between anchor joints which captures the adjacency configuration in the skeletal pattern. Thus, the kinematic relationship amongst the body joints across the frames using pose estimation helps in better establishing the spatio-temporal dependencies. We present an exhaustive empirical evaluation of state-of-the-art deep network based methods for dance classification on ICD dataset.

Prerana Mukherjee, Brejesh Lall, Snehith Lattupally

Object cosegmentation addresses the problem of discovering similar objects from multiple images and segmenting them as foreground simultaneously. In this paper, we propose a novel end-to-end pipeline to segment the similar objects simultaneously from relevant set of images using supervised learning via deep-learning framework. We experiment with multiple set of object proposal generation techniques and perform extensive numerical evaluations by training the Siamese network with generated object proposals. Similar objects proposals for the test images are retrieved using the ANNOY (Approximate Nearest Neighbor) library and deep semantic segmentation is performed on them. Finally, we form a collage from the segmented similar objects based on the relative importance of the objects.

Aarushi Agrawal, Prerana Mukherjee, Siddharth Srivastava et al.

Text spotting is an interesting research problem as text may appear at any random place and may occur in various forms. Moreover, ability to detect text opens the horizons for improving many advanced computer vision problems. In this paper, we propose a novel language agnostic text detection method utilizing edge enhanced Maximally Stable Extremal Regions in natural scenes by defining strong characterness measures. We show that a simple combination of characterness cues help in rejecting the non text regions. These regions are further fine-tuned for rejecting the non-textual neighbor regions. Comprehensive evaluation of the proposed scheme shows that it provides comparative to better generalization performance to the traditional methods for this task.