Vinay P. Namboodiri

Badri N. Patro, Vinay P. Namboodiri, Vijay Srinivas Agneeswaran

Vision transformers have been applied successfully for image recognition tasks. There have been either multi-headed self-attention based (ViT \cite{dosovitskiy2020image}, DeIT, \cite{touvron2021training}) similar to the original work in textual models or more recently based on spectral layers (Fnet\cite{lee2021fnet}, GFNet\cite{rao2021global}, AFNO\cite{guibas2021efficient}). We hypothesize that both spectral and multi-headed attention plays a major role. We investigate this hypothesis through this work and observe that indeed combining spectral and multi-headed attention layers provides a better transformer architecture. We thus propose the novel Spectformer architecture for transformers that combines spectral and multi-headed attention layers. We believe that the resulting representation allows the transformer to capture the feature representation appropriately and it yields improved performance over other transformer representations. For instance, it improves the top-1 accuracy by 2\% on ImageNet compared to both GFNet-H and LiT. SpectFormer-S reaches 84.25\% top-1 accuracy on ImageNet-1K (state of the art for small version). Further, Spectformer-L achieves 85.7\% that is the state of the art for the comparable base version of the transformers. We further ensure that we obtain reasonable results in other scenarios such as transfer learning on standard datasets such as CIFAR-10, CIFAR-100, Oxford-IIIT-flower, and Standford Car datasets. We then investigate its use in downstream tasks such of object detection and instance segmentation on the MS-COCO dataset and observe that Spectformer shows consistent performance that is comparable to the best backbones and can be further optimized and improved. Hence, we believe that combined spectral and attention layers are what are needed for vision transformers.

Madhav Agarwal, Anchit Gupta, Rudrabha Mukhopadhyay et al.

We leverage the modern advancements in talking head generation to propose an end-to-end system for talking head video compression. Our algorithm transmits pivot frames intermittently while the rest of the talking head video is generated by animating them. We use a state-of-the-art face reenactment network to detect key points in the non-pivot frames and transmit them to the receiver. A dense flow is then calculated to warp a pivot frame to reconstruct the non-pivot ones. Transmitting key points instead of full frames leads to significant compression. We propose a novel algorithm to adaptively select the best-suited pivot frames at regular intervals to provide a smooth experience. We also propose a frame-interpolater at the receiver's end to improve the compression levels further. Finally, a face enhancement network improves reconstruction quality, significantly improving several aspects like the sharpness of the generations. We evaluate our method both qualitatively and quantitatively on benchmark datasets and compare it with multiple compression techniques. We release a demo video and additional information at https://cvit.iiit.ac.in/research/projects/cvit-projects/talking-video-compression.

Munender Varshney, Ravindra Yadav, Vinay P. Namboodiri et al.

Understanding the lip movement and inferring the speech from it is notoriously difficult for the common person. The task of accurate lip-reading gets help from various cues of the speaker and its contextual or environmental setting. Every speaker has a different accent and speaking style, which can be inferred from their visual and speech features. This work aims to understand the correlation/mapping between speech and the sequence of lip movement of individual speakers in an unconstrained and large vocabulary. We model the frame sequence as a prior to the transformer in an auto-encoder setting and learned a joint embedding that exploits temporal properties of both audio and video. We learn temporal synchronization using deep metric learning, which guides the decoder to generate speech in sync with input lip movements. The predictive posterior thus gives us the generated speech in speaker speaking style. We have trained our model on the Grid and Lip2Wav Chemistry lecture dataset to evaluate single speaker natural speech generation tasks from lip movement in an unconstrained natural setting. Extensive evaluation using various qualitative and quantitative metrics with human evaluation also shows that our method outperforms the Lip2Wav Chemistry dataset(large vocabulary in an unconstrained setting) by a good margin across almost all evaluation metrics and marginally outperforms the state-of-the-art on GRID dataset.

Soumya Banerjee, Vinay K. Verma, Avideep Mukherjee et al.

Lifelong learning or continual learning is the problem of training an AI agent continuously while also preventing it from forgetting its previously acquired knowledge. Streaming lifelong learning is a challenging setting of lifelong learning with the goal of continuous learning in a dynamic non-stationary environment without forgetting. We introduce a novel approach to lifelong learning, which is streaming (observes each training example only once), requires a single pass over the data, can learn in a class-incremental manner, and can be evaluated on-the-fly (anytime inference). To accomplish these, we propose a novel \emph{virtual gradients} based approach for continual representation learning which adapts to each new example while also generalizing well on past data to prevent catastrophic forgetting. Our approach also leverages an exponential-moving-average-based semantic memory to further enhance performance. Experiments on diverse datasets with temporally correlated observations demonstrate our method's efficacy and superior performance over existing methods.

Utsav Singh, Vinay P. Namboodiri

Hierarchical reinforcement learning (HRL) has the potential to solve complex long horizon tasks using temporal abstraction and increased exploration. However, hierarchical agents are difficult to train due to inherent non-stationarity. We present primitive enabled adaptive relabeling (PEAR), a two-phase approach where we first perform adaptive relabeling on a few expert demonstrations to generate efficient subgoal supervision, and then jointly optimize HRL agents by employing reinforcement learning (RL) and imitation learning (IL). We perform theoretical analysis to bound the sub-optimality of our approach and derive a joint optimization framework using RL and IL. Since PEAR utilizes only a few expert demonstrations and considers minimal limiting assumptions on the task structure, it can be easily integrated with typical off-policy RL algorithms to produce a practical HRL approach. We perform extensive experiments on challenging environments and show that PEAR is able to outperform various hierarchical and non-hierarchical baselines and achieve upto $80\%$ success rates in complex sparse robotic control tasks where other baselines typically fail to show significant progress. We also perform ablations to thoroughly analyse the importance of our various design choices. Finally, we perform real world robotic experiments on complex tasks and demonstrate that PEAR consistently outperforms the baselines.

Cangxiong Chen, Vinay P. Namboodiri, Julian Padget

CLIP is a widely used foundational vision-language model that is used for zero-shot image recognition and other image-text alignment tasks. We demonstrate that CLIP is vulnerable to change in image quality under compression. This surprising result is further analysed using an attribution method-Integrated Gradients. Using this attribution method, we are able to better understand both quantitatively and qualitatively exactly the nature in which the compression affects the zero-shot recognition accuracy of this model. We evaluate this extensively on CIFAR-10 and STL-10. Our work provides the basis to understand this vulnerability of CLIP and can help us develop more effective methods to improve the robustness of CLIP and other vision-language models.

Soumya Banerjee, Vinay Kumar Verma, Vinay P. Namboodiri

Despite rapid advancements in lifelong learning (LLL) research, a large body of research mainly focuses on improving the performance in the existing \textit{static} continual learning (CL) setups. These methods lack the ability to succeed in a rapidly changing \textit{dynamic} environment, where an AI agent needs to quickly learn new instances in a `single pass' from the non-i.i.d (also possibly temporally contiguous/coherent) data streams without suffering from catastrophic forgetting. For practical applicability, we propose a novel lifelong learning approach, which is streaming, i.e., a single input sample arrives in each time step, single pass, class-incremental, and subject to be evaluated at any moment. To address this challenging setup and various evaluation protocols, we propose a Bayesian framework, that enables fast parameter update, given a single training example, and enables any-time inference. We additionally propose an implicit regularizer in the form of snap-shot self-distillation, which effectively minimizes the forgetting further. We further propose an effective method that efficiently selects a subset of samples for online memory rehearsal and employs a new replay buffer management scheme that significantly boosts the overall performance. Our empirical evaluations and ablations demonstrate that the proposed method outperforms the prior works by large margins.

Avideep Mukherjee, Soumya Banerjee, Piyush Rai et al.

Diffusion-based models demonstrate impressive generation capabilities. However, they also have a massive number of parameters, resulting in enormous model sizes, thus making them unsuitable for deployment on resource-constraint devices. Block-wise generation can be a promising alternative for designing compact-sized (parameter-efficient) deep generative models since the model can generate one block at a time instead of generating the whole image at once. However, block-wise generation is also considerably challenging because ensuring coherence across generated blocks can be non-trivial. To this end, we design a retrieval-augmented generation (RAG) approach and leverage the corresponding blocks of the images retrieved by the RAG module to condition the training and generation stages of a block-wise denoising diffusion model. Our conditioning schemes ensure coherence across the different blocks during training and, consequently, during generation. While we showcase our approach using the latent diffusion model (LDM) as the base model, it can be used with other variants of denoising diffusion models. We validate the solution of the coherence problem through the proposed approach by reporting substantive experiments to demonstrate our approach's effectiveness in compact model size and excellent generation quality.

Samik Some, Vinay P. Namboodiri

Present-day deep neural networks for video semantic segmentation require a large number of fine-grained pixel-level annotations to achieve the best possible results. Obtaining such annotations, however, is very expensive. On the other hand, raw, unannotated video frames are practically free to obtain. Similarly, coarse annotations, which do not require precise boundaries, are also much cheaper. This paper investigates approaches to reduce the annotation cost required for video segmentation datasets by utilising such resources. We show that using state-of-the-art segmentation foundation models, Segment Anything Model (SAM) and Segment Anything Model 2 (SAM 2), we can utilise both unannotated frames as well as coarse annotations to alleviate the effort required for manual annotation of video segmentation datasets by automating mask generation. Our investigation suggests that if used appropriately, we can reduce the need for annotation by a third with similar performance for video semantic segmentation. More significantly, our analysis suggests that the variety of frames in the dataset is more important than the number of frames for obtaining the best performance.

Alexey Kravets, Da Li, Chuan Li et al.

Recent advances in language model interpretability using sparse autoencoders (SAEs) have yet to effectively translate to the visual domain, mainly due to the difficulty and ambiguity of labeling visual concepts. In this paper, we introduce Visual Interpretability via SAE Transfer Alignment (VISTA), a framework that transfers interpretability from language to vision in a LLaVA-style vision-language model by constraining a visual projector to map visual tokens into an LLM's pre-existing, labeled textual SAE space. This approach enables visual interpretability without training dedicated vision SAEs. By regularizing the projector using the LLM's SAE reconstruction loss, VISTA achieves a threefold increase in the matching rate, which measures how accurately the most activating textual concepts in the SAE space correspond to semantic elements in the image. Using this framework, we further analyze spatial localization properties of different vision encoders and show that DINOv2 features have stronger localization abilities than other encoders. Leveraging this precision, we validate VISTA's cross-modal alignment through fine-grained, localized concept interventions, where specific objects are removed or replaced in the model's perception while preserving the surrounding scene. This results in improvements of 35% in object removal and 47% in object replacement tasks over vision-only baselines, providing causal evidence that visual tokens inhabit the text SAE manifold. These contributions are validated across multiple LLM architectures.

Badri N. Patro, Suhas Ranganath, Vinay P. Namboodiri et al.

Transformers have revolutionized image modeling tasks with adaptations like DeIT, Swin, SVT, Biformer, STVit, and FDVIT. However, these models often face challenges with inductive bias and high quadratic complexity, making them less efficient for high-resolution images. State space models (SSMs) such as Mamba, V-Mamba, ViM, and SiMBA offer an alternative to handle high resolution images in computer vision tasks. These SSMs encounter two major issues. First, they become unstable when scaled to large network sizes. Second, although they efficiently capture global information in images, they inherently struggle with handling local information. To address these challenges, we introduce Heracles, a novel SSM that integrates a local SSM, a global SSM, and an attention-based token interaction module. Heracles leverages a Hartely kernel-based state space model for global image information, a localized convolutional network for local details, and attention mechanisms in deeper layers for token interactions. Our extensive experiments demonstrate that Heracles-C-small achieves state-of-the-art performance on the ImageNet dataset with 84.5\% top-1 accuracy. Heracles-C-Large and Heracles-C-Huge further improve accuracy to 85.9\% and 86.4\%, respectively. Additionally, Heracles excels in transfer learning tasks on datasets such as CIFAR-10, CIFAR-100, Oxford Flowers, and Stanford Cars, and in instance segmentation on the MSCOCO dataset. Heracles also proves its versatility by achieving state-of-the-art results on seven time-series datasets, showcasing its ability to generalize across domains with spectral data, capturing both local and global information. The project page is available at this link.\url{https://github.com/badripatro/heracles}

Badri N. Patro, Shivansh Pate, Vinay P. Namboodiri

In this paper, we propose a method to obtain robust explanations for visual question answering(VQA) that correlate well with the answers. Our model explains the answers obtained through a VQA model by providing visual and textual explanations. The main challenges that we address are i) Answers and textual explanations obtained by current methods are not well correlated and ii) Current methods for visual explanation do not focus on the right location for explaining the answer. We address both these challenges by using a collaborative correlated module which ensures that even if we do not train for noise based attacks, the enhanced correlation ensures that the right explanation and answer can be generated. We further show that this also aids in improving the generated visual and textual explanations. The use of the correlated module can be thought of as a robust method to verify if the answer and explanations are coherent. We evaluate this model using VQA-X dataset. We observe that the proposed method yields better textual and visual justification that supports the decision. We showcase the robustness of the model against a noise-based perturbation attack using corresponding visual and textual explanations. A detailed empirical analysis is shown. Here we provide source code link for our model \url{https://github.com/DelTA-Lab-IITK/CCM-WACV}.

Saurabh Srivastava, Vinay P. Namboodiri, T. V. Prabhakar

AI intensive systems that operate upon user data face the challenge of balancing data utility with privacy concerns. We propose the idea and present the prototype of an open-source tool called Privacy Utility Trade-off (PUT) Workbench which seeks to aid software practitioners to take such crucial decisions. We pick a simple privacy model that doesn't require any background knowledge in Data Science and show how even that can achieve significant results over standard and real-life datasets. The tool and the source code is made freely available for extensions and usage.

Utsav Singh, Vinay P. Namboodiri

Hierarchical reinforcement learning (HRL) leverages temporal abstraction to efficiently tackle complex long-horizon tasks. However, HRL often collapses because the continual updates of the low-level primitive make earlier sub-goals issued by the high-level policy obsolete, introducing non-stationarity that destabilizes training. We propose CRISP, a curriculum-driven framework that tackles this instability with three key ingredients: (1) primitive-informed parsing (PIP), which adaptively re-labels a handful of expert demonstrations to always generate reachable subgoals by the current low-level primitive, (2) an inverse-reinforcement-learning regularizer that steers the high-level policy toward the expert-induced subgoal distribution and stabilizes learning, and (3) a unified training loop that leverages these components to boost sample efficiency. Across six sparse-reward robotic navigation and manipulation benchmarks, CRISP improves success rates by more than 40% over strong hierarchical and flat baselines and successfully transfers to real-world tasks, demonstrating the promise of curriculum-based HRL for practical scenarios.

Utsav Singh, Wesley A. Suttle, Brian M. Sadler et al.

In this work, we introduce PIPER: Primitive-Informed Preference-based Hierarchical reinforcement learning via Hindsight Relabeling, a novel approach that leverages preference-based learning to learn a reward model, and subsequently uses this reward model to relabel higher-level replay buffers. Since this reward is unaffected by lower primitive behavior, our relabeling-based approach is able to mitigate non-stationarity, which is common in existing hierarchical approaches, and demonstrates impressive performance across a range of challenging sparse-reward tasks. Since obtaining human feedback is typically impractical, we propose to replace the human-in-the-loop approach with our primitive-in-the-loop approach, which generates feedback using sparse rewards provided by the environment. Moreover, in order to prevent infeasible subgoal prediction and avoid degenerate solutions, we propose primitive-informed regularization that conditions higher-level policies to generate feasible subgoals for lower-level policies. We perform extensive experiments to show that PIPER mitigates non-stationarity in hierarchical reinforcement learning and achieves greater than 50$\%$ success rates in challenging, sparse-reward robotic environments, where most other baselines fail to achieve any significant progress.

Jack Saunders, Charlie Hewitt, Yanan Jian et al.

Gaussian Splatting has changed the game for real-time photo-realistic rendering. One of the most popular applications of Gaussian Splatting is to create animatable avatars, known as Gaussian Avatars. Recent works have pushed the boundaries of quality and rendering efficiency but suffer from two main limitations. Either they require expensive multi-camera rigs to produce avatars with free-view rendering, or they can be trained with a single camera but only rendered at high quality from this fixed viewpoint. An ideal model would be trained using a short monocular video or image from available hardware, such as a webcam, and rendered from any view. To this end, we propose GASP: Gaussian Avatars with Synthetic Priors. To overcome the limitations of existing datasets, we exploit the pixel-perfect nature of synthetic data to train a Gaussian Avatar prior. By fitting this prior model to a single photo or video and fine-tuning it, we get a high-quality Gaussian Avatar, which supports 360$^\circ$ rendering. Our prior is only required for fitting, not inference, enabling real-time application. Through our method, we obtain high-quality, animatable Avatars from limited data which can be animated and rendered at 70fps on commercial hardware. See our project page (https://microsoft.github.io/GASP/) for results.

Utsav Singh, Souradip Chakraborty, Wesley A. Suttle et al.

Hierarchical reinforcement learning (HRL) enables agents to solve complex, long-horizon tasks by decomposing them into manageable sub-tasks. However, HRL methods often suffer from two fundamental challenges: (i) non-stationarity, caused by the changing behavior of the lower-level policy during training, which destabilizes higher-level policy learning, and (ii) the generation of infeasible subgoals that lower-level policies cannot achieve. In this work, we introduce DIPPER, a novel HRL framework that formulates hierarchical policy learning as a bi-level optimization problem and leverages direct preference optimization (DPO) to train the higher-level policy using preference feedback. By optimizing the higher-level policy with DPO, we decouple higher-level learning from the non-stationary lower-level reward signal, thus mitigating non-stationarity. To further address the infeasible subgoal problem, DIPPER incorporates a regularization that tries to ensure the feasibility of subgoal tasks within the capabilities of the lower-level policy. Extensive experiments on challenging robotic navigation and manipulation benchmarks demonstrate that DIPPER achieves up to 40\% improvement over state-of-the-art baselines in sparse reward scenarios, highlighting its effectiveness in overcoming longstanding limitations of HRL.

Alexey Kravets, Da Chen, Vinay P. Namboodiri

CLIP is a foundational model with transferable classification performance in the few-shot setting. Several methods have shown improved performance of CLIP using few-shot examples. However, so far, all these techniques have been benchmarked using standard few-shot datasets. We argue that this mode of evaluation does not provide a true indication of the inductive generalization ability using few-shot examples. As most datasets have been seen by the CLIP model, the resultant setting can be termed as partially transductive. To solve this, we propose a pipeline that uses an unlearning technique to obtain true inductive baselines. In this new inductive setting, the methods show a significant drop in performance (-55% on average among 13 baselines with multiple datasets). We validate the unlearning technique using oracle baselines. An improved few-shot classification technique is proposed that consistently obtains state-of-the-art performance over 13 other recent baseline methods on a comprehensive analysis with 5880 experiments - varying the datasets, differing number of few-shot examples, unlearning setting, and with different seeds. Thus, we identify the issue with the evaluation of CLIP-based few-shot classification, provide a solution using unlearning, propose new benchmarks, and provide an improved method.

Diljeet Jagpal, Xi Chen, Vinay P. Namboodiri

Zero-shot, training-free, image-based text-to-video generation is an emerging area that aims to generate videos using existing image-based diffusion models. Current methods in this space require specific architectural changes to image generation models, which limit their adaptability and scalability. In contrast to such methods, we provide a model-agnostic approach. We use intersections in diffusion trajectories, working only with the latent values. We could not obtain localized frame-wise coherence and diversity using only the intersection of trajectories. Thus, we instead use a grid-based approach. An in-context trained LLM is used to generate coherent frame-wise prompts; another is used to identify differences between frames. Based on these, we obtain a CLIP-based attention mask that controls the timing of switching the prompts for each grid cell. Earlier switching results in higher variance, while later switching results in more coherence. Therefore, our approach can ensure appropriate control between coherence and variance for the frames. Our approach results in state-of-the-art performance while being more flexible when working with diverse image-generation models. The empirical analysis using quantitative metrics and user studies confirms our model's superior temporal consistency, visual fidelity and user satisfaction, thus providing a novel way to obtain training-free, image-based text-to-video generation.

Cangxiong Chen, Vinay P. Namboodiri, Julia E. Sero

The spatio-temporal nature of live-cell microscopy data poses challenges in the analysis of cell states which is fundamental in bioimaging. Deep-learning based segmentation or tracking methods rely on large amount of high quality annotations to work effectively. In this work, we explore an alternative solution: using feature maps obtained from self-supervised representation learning (SSRL) on time arrow prediction (TAP) for the downstream supervised task of cell event recognition. We demonstrate through extensive experiments and analysis that this approach can achieve better performance with limited annotation compared to models trained from end to end using fully supervised approach. Our analysis also provides insight into applications of the SSRL using TAP in live-cell microscopy.

Samik Some, Vinay P. Namboodiri

Semantic segmentation has become an important task in computer vision with the growth of self-driving cars, medical image segmentation, etc. Although current models provide excellent results, they are still far from perfect and while there has been significant work in trying to improve the performance, both with respect to accuracy and speed of segmentation, there has been little work which analyses the failure cases of such systems. In this work, we aim to provide an analysis of how segmentation fails across different models and consider the question of whether these can be predicted reasonably at test time. To do so, we explore existing uncertainty-based metrics and see how well they correlate with misclassifications, allowing us to define the degree of trust we put in the output of our prediction models. Through several experiments on three different models across three datasets, we show that simple measures such as entropy can be used to capture misclassification with high recall rates.

Utsav Singh, Pramit Bhattacharyya, Vinay P. Namboodiri

Large language models (LLMs) have shown remarkable abilities in logical reasoning, in-context learning, and code generation. However, translating natural language instructions into effective robotic control policies remains a significant challenge, especially for tasks requiring long-horizon planning and operating under sparse reward conditions. Hierarchical Reinforcement Learning (HRL) provides a natural framework to address this challenge in robotics; however, it typically suffers from non-stationarity caused by the changing behavior of the lower-level policy during training, destabilizing higher-level policy learning. We introduce LGR2, a novel HRL framework that leverages LLMs to generate language-guided reward functions for the higher-level policy. By decoupling high-level reward generation from low-level policy changes, LGR2 fundamentally mitigates the non-stationarity problem in off-policy HRL, enabling stable and efficient learning. To further enhance sample efficiency in sparse environments, we integrate goal-conditioned hindsight experience relabeling. Extensive experiments across simulated and real-world robotic navigation and manipulation tasks demonstrate LGR2 outperforms both hierarchical and non-hierarchical baselines, achieving over 55% success rates on challenging tasks and robust transfer to real robots, without additional fine-tuning.

Vinod K Kurmi, Rishabh Sharma, Yash Vardhan Sharma et al.

Bias mitigation in machine learning models is imperative, yet challenging. While several approaches have been proposed, one view towards mitigating bias is through adversarial learning. A discriminator is used to identify the bias attributes such as gender, age or race in question. This discriminator is used adversarially to ensure that it cannot distinguish the bias attributes. The main drawback in such a model is that it directly introduces a trade-off with accuracy as the features that the discriminator deems to be sensitive for discrimination of bias could be correlated with classification. In this work we solve the problem. We show that a biased discriminator can actually be used to improve this bias-accuracy tradeoff. Specifically, this is achieved by using a feature masking approach using the discriminator's gradients. We ensure that the features favoured for the bias discrimination are de-emphasized and the unbiased features are enhanced during classification. We show that this simple approach works well to reduce bias as well as improve accuracy significantly. We evaluate the proposed model on standard benchmarks. We improve the accuracy of the adversarial methods while maintaining or even improving the unbiasness and also outperform several other recent methods.

Soumya Banerjee, Vinay Kumar Verma, Toufiq Parag et al.

A wide variety of methods have been developed to enable lifelong learning in conventional deep neural networks. However, to succeed, these methods require a `batch' of samples to be available and visited multiple times during training. While this works well in a static setting, these methods continue to suffer in a more realistic situation where data arrives in \emph{online streaming manner}. We empirically demonstrate that the performance of current approaches degrades if the input is obtained as a stream of data with the following restrictions: $(i)$ each instance comes one at a time and can be seen only once, and $(ii)$ the input data violates the i.i.d assumption, i.e., there can be a class-based correlation. We propose a novel approach (CIOSL) for the class-incremental learning in an \emph{online streaming setting} to address these challenges. The proposed approach leverages implicit and explicit dual weight regularization and experience replay. The implicit regularization is leveraged via the knowledge distillation, while the explicit regularization incorporates a novel approach for parameter regularization by learning the joint distribution of the buffer replay and the current sample. Also, we propose an efficient online memory replay and replacement buffer strategy that significantly boosts the model's performance. Extensive experiments and ablation on challenging datasets show the efficacy of the proposed method.

Anchit Gupta, Faizan Farooq Khan, Rudrabha Mukhopadhyay et al.

This paper proposes a video editor based on OpenShot with several state-of-the-art facial video editing algorithms as added functionalities. Our editor provides an easy-to-use interface to apply modern lip-syncing algorithms interactively. Apart from lip-syncing, the editor also uses audio and facial re-enactment to generate expressive talking faces. The manual control improves the overall experience of video editing without missing out on the benefits of modern synthetic video generation algorithms. This control enables us to lip-sync complex dubbed movie scenes, interviews, television shows, and other visual content. Furthermore, our editor provides features that automatically translate lectures from spoken content, lip-sync of the professor, and background content like slides. While doing so, we also tackle the critical aspect of synchronizing background content with the translated speech. We qualitatively evaluate the usefulness of the proposed editor by conducting human evaluations. Our evaluations show a clear improvement in the efficiency of using human editors and an improved video generation quality. We attach demo videos with the supplementary material clearly explaining the tool and also showcasing multiple results.

Avideep Mukherjee, Badri Narayan Patro, Vinay P. Namboodiri

Normalizing flows provide an elegant method for obtaining tractable density estimates from distributions by using invertible transformations. The main challenge is to improve the expressivity of the models while keeping the invertibility constraints intact. We propose to do so via the incorporation of localized self-attention. However, conventional self-attention mechanisms don't satisfy the requirements to obtain invertible flows and can't be naively incorporated into normalizing flows. To address this, we introduce a novel approach called Attentive Contractive Flow (ACF) which utilizes a special category of flow-based generative models - contractive flows. We demonstrate that ACF can be introduced into a variety of state of the art flow models in a plug-and-play manner. This is demonstrated to not only improve the representation power of these models (improving on the bits per dim metric), but also to results in significantly faster convergence in training them. Qualitative results, including interpolations between test images, demonstrate that samples are more realistic and capture local correlations in the data well. We evaluate the results further by performing perturbation analysis using AWGN demonstrating that ACF models (especially the dot-product variant) show better and more consistent resilience to additive noise.

Zeeshan Khan, Kartheek Akella, Vinay P. Namboodiri et al.

This work studies the long-standing problems of model capacity and negative interference in multilingual neural machine translation MNMT. We use network pruning techniques and observe that pruning 50-70% of the parameters from a trained MNMT model results only in a 0.29-1.98 drop in the BLEU score. Suggesting that there exist large redundancies even in MNMT models. These observations motivate us to use the redundant parameters and counter the interference problem efficiently. We propose a novel adaptation strategy, where we iteratively prune and retrain the redundant parameters of an MNMT to improve bilingual representations while retaining the multilinguality. Negative interference severely affects high resource languages, and our method alleviates it without any additional adapter modules. Hence, we call it parameter-free adaptation strategy, paving way for the efficient adaptation of MNMT. We demonstrate the effectiveness of our method on a 9 language MNMT trained on TED talks, and report an average improvement of +1.36 BLEU on high resource pairs. Code will be released here.

Blessen George, Vinod K. Kurmi, Vinay P. Namboodiri

Generative adversarial networks (GANs) are very popular to generate realistic images, but they often suffer from the training instability issues and the phenomenon of mode loss. In order to attain greater diversity in GAN synthesized data, it is critical to solving the problem of mode loss. Our work explores probabilistic approaches to GAN modelling that could allow us to tackle these issues. We present Prb-GANs, a new variation that uses dropout to create a distribution over the network parameters with the posterior learnt using variational inference. We describe theoretically and validate experimentally using simple and complex datasets the benefits of such an approach. We look into further improvements using the concept of uncertainty measures. Through a set of further modifications to the loss functions for each network of the GAN, we are able to get results that show the improvement of GAN performance. Our methods are extremely simple and require very little modification to existing GAN architecture.

Vinod K Kurmi, Venkatesh K Subramanian, Vinay P. Namboodiri

Adaptation of a classifier to new domains is one of the challenging problems in machine learning. This has been addressed using many deep and non-deep learning based methods. Among the methodologies used, that of adversarial learning is widely applied to solve many deep learning problems along with domain adaptation. These methods are based on a discriminator that ensures source and target distributions are close. However, here we suggest that rather than using a point estimate obtaining by a single discriminator, it would be useful if a distribution based on ensembles of discriminators could be used to bridge this gap. This could be achieved using multiple classifiers or using traditional ensemble methods. In contrast, we suggest that a Monte Carlo dropout based ensemble discriminator could suffice to obtain the distribution based discriminator. Specifically, we propose a curriculum based dropout discriminator that gradually increases the variance of the sample based distribution and the corresponding reverse gradients are used to align the source and target feature representations. An ensemble of discriminators helps the model to learn the data distribution efficiently. It also provides a better gradient estimates to train the feature extractor. The detailed results and thorough ablation analysis show that our model outperforms state-of-the-art results.

Pratik Mazumder, Pravendra Singh, Vinay P. Namboodiri

Deep learning models generally learn the biases present in the training data. Researchers have proposed several approaches to mitigate such biases and make the model fair. Bias mitigation techniques assume that a sufficiently large number of training examples are present. However, we observe that if the training data is limited, then the effectiveness of bias mitigation methods is severely degraded. In this paper, we propose a novel approach to address this problem. Specifically, we adapt self-supervision and self-distillation to reduce the impact of biases on the model in this setting. Self-supervision and self-distillation are not used for bias mitigation. However, through this work, we demonstrate for the first time that these techniques are very effective in bias mitigation. We empirically show that our approach can significantly reduce the biases learned by the model. Further, we experimentally demonstrate that our approach is complementary to other bias mitigation strategies. Our approach significantly improves their performance and further reduces the model biases in the limited data regime. Specifically, on the L-CIFAR-10S skewed dataset, our approach significantly reduces the bias score of the baseline model by 78.22% and outperforms it in terms of accuracy by a significant absolute margin of 8.89%. It also significantly reduces the bias score for the state-of-the-art domain independent bias mitigation method by 59.26% and improves its performance by a significant absolute margin of 7.08%.

Pravendra Singh, Pratik Mazumder, Piyush Rai et al.

Deep learning models suffer from catastrophic forgetting when trained in an incremental learning setting. In this work, we propose a novel approach to address the task incremental learning problem, which involves training a model on new tasks that arrive in an incremental manner. The task incremental learning problem becomes even more challenging when the test set contains classes that are not part of the train set, i.e., a task incremental generalized zero-shot learning problem. Our approach can be used in both the zero-shot and non zero-shot task incremental learning settings. Our proposed method uses weight rectifications and affine transformations in order to adapt the model to different tasks that arrive sequentially. Specifically, we adapt the network weights to work for new tasks by "rectifying" the weights learned from the previous task. We learn these weight rectifications using very few parameters. We additionally learn affine transformations on the outputs generated by the network in order to better adapt them for the new task. We perform experiments on several datasets in both zero-shot and non zero-shot task incremental learning settings and empirically show that our approach achieves state-of-the-art results. Specifically, our approach outperforms the state-of-the-art non zero-shot task incremental learning method by over 5% on the CIFAR-100 dataset. Our approach also significantly outperforms the state-of-the-art task incremental generalized zero-shot learning method by absolute margins of 6.91% and 6.33% for the AWA1 and CUB datasets, respectively. We validate our approach using various ablation studies.

Vinod K Kurmi, Badri N. Patro, Venkatesh K. Subramanian et al.

One of the major limitations of deep learning models is that they face catastrophic forgetting in an incremental learning scenario. There have been several approaches proposed to tackle the problem of incremental learning. Most of these methods are based on knowledge distillation and do not adequately utilize the information provided by older task models, such as uncertainty estimation in predictions. The predictive uncertainty provides the distributional information can be applied to mitigate catastrophic forgetting in a deep learning framework. In the proposed work, we consider a Bayesian formulation to obtain the data and model uncertainties. We also incorporate self-attention framework to address the incremental learning problem. We define distillation losses in terms of aleatoric uncertainty and self-attention. In the proposed work, we investigate different ablation analyses on these losses. Furthermore, we are able to obtain better results in terms of accuracy on standard benchmarks.

Kartheek Akella, Sai Himal Allu, Sridhar Suresh Ragupathi et al.

In this paper, we address the task of improving pair-wise machine translation for specific low resource Indian languages. Multilingual NMT models have demonstrated a reasonable amount of effectiveness on resource-poor languages. In this work, we show that the performance of these models can be significantly improved upon by using back-translation through a filtered back-translation process and subsequent fine-tuning on the limited pair-wise language corpora. The analysis in this paper suggests that this method can significantly improve a multilingual model's performance over its baseline, yielding state-of-the-art results for various Indian languages.

Pratik Mazumder, Pravendra Singh, Vinay P. Namboodiri

Learning from a few examples is an important practical aspect of training classifiers. Various works have examined this aspect quite well. However, all existing approaches assume that the few examples provided are always correctly labeled. This is a strong assumption, especially if one considers the current techniques for labeling using crowd-based labeling services. We address this issue by proposing a novel robust few-shot learning approach. Our method relies on generating robust prototypes from a set of few examples. Specifically, our method refines the class prototypes by producing hybrid features from the support examples of each class. The refined prototypes help to classify the query images better. Our method can replace the evaluation phase of any few-shot learning method that uses a nearest neighbor prototype-based evaluation procedure to make them robust. We evaluate our method on standard mini-ImageNet and tiered-ImageNet datasets. We perform experiments with various label corruption rates in the support examples of the few-shot classes. We obtain significant improvement over widely used few-shot learning methods that suffer significant performance degeneration in the presence of label noise. We finally provide extensive ablation experiments to validate our method.

Raghav B. Venkataramaiyer, Abhishek Joshi, Saisha Narang et al.

Hatching is a common method used by artists to accentuate the third dimension of a sketch, and to illuminate the scene. Our system SHAD3S attempts to compete with a human at hatching generic three-dimensional (3D) shapes, and also tries to assist her in a form exploration exercise. The novelty of our approach lies in the fact that we make no assumptions about the input other than that it represents a 3D shape, and yet, given a contextual information of illumination and texture, we synthesise an accurate hatch pattern over the sketch, without access to 3D or pseudo 3D. In the process, we contribute towards a) a cheap yet effective method to synthesise a sufficiently large high fidelity dataset, pertinent to task; b) creating a pipeline with conditional generative adversarial network (CGAN); and c) creating an interactive utility with GIMP, that is a tool for artists to engage with automated hatching or a form-exploration exercise. User evaluation of the tool suggests that the model performance does generalise satisfactorily over diverse input, both in terms of style as well as shape. A simple comparison of inception scores suggest that the generated distribution is as diverse as the ground truth.

Samik Some, Mithun Das Gupta, Vinay P. Namboodiri

We present a determinantal point process (DPP) inspired alternative to non-maximum suppression (NMS) which has become an integral step in all state-of-the-art object detection frameworks. DPPs have been shown to encourage diversity in subset selection problems. We pose NMS as a subset selection problem and posit that directly incorporating DPP like framework can improve the overall performance of the object detection system. We propose an optimization problem which takes the same inputs as NMS, but introduces a novel sub-modularity based diverse subset selection functional. Our results strongly indicate that the modifications proposed in this paper can provide consistent improvements to state-of-the-art object detection pipelines.

Jerin Philip, Shashank Siripragada, Vinay P. Namboodiri et al.

Indian language machine translation performance is hampered due to the lack of large scale multi-lingual sentence aligned corpora and robust benchmarks. Through this paper, we provide and analyse an automated framework to obtain such a corpus for Indian language neural machine translation (NMT) systems. Our pipeline consists of a baseline NMT system, a retrieval module, and an alignment module that is used to work with publicly available websites such as press releases by the government. The main contribution towards this effort is to obtain an incremental method that uses the above pipeline to iteratively improve the size of the corpus as well as improve each of the components of our system. Through our work, we also evaluate the design choices such as the choice of pivoting language and the effect of iterative incremental increase in corpus size. Our work in addition to providing an automated framework also results in generating a relatively larger corpus as compared to existing corpora that are available for Indian languages. This corpus helps us obtain substantially improved results on the publicly available WAT evaluation benchmark and other standard evaluation benchmarks.

Shashank Siripragada, Jerin Philip, Vinay P. Namboodiri et al.

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.

Prem Raj, Vinay P. Namboodiri, L. Behera

Convolutional Neural Networks (CNNs) have been successfully applied for relative camera pose estimation from labeled image-pair data, without requiring any hand-engineered features, camera intrinsic parameters or depth information. The trained CNN can be utilized for performing pose based visual servo control (PBVS). One of the ways to improve the quality of visual servo output is to improve the accuracy of the CNN for estimating the relative pose estimation. With a given state-of-the-art CNN for relative pose regression, how can we achieve an improved performance for visual servo control? In this paper, we explore switching of CNNs to improve the precision of visual servo control. The idea of switching a CNN is due to the fact that the dataset for training a relative camera pose regressor for visual servo control must contain variations in relative pose ranging from a very small scale to eventually a larger scale. We found that, training two different instances of the CNN, one for large-scale-displacements (LSD) and another for small-scale-displacements (SSD) and switching them during the visual servo execution yields better results than training a single CNN with the combined LSD+SSD data. However, it causes extra storage overhead and switching decision is taken by a manually set threshold which may not be optimal for all the scenes. To eliminate these drawbacks, we propose an efficient switching strategy based on model agnostic meta learning (MAML) algorithm. In this, a single model is trained to learn parameters which are simultaneously good for multiple tasks, namely a binary classification for switching decision, a 6DOF pose regression for LSD data and also a 6DOF pose regression for SSD data. The proposed approach performs far better than the naive approach, while storage and run-time overheads are almost negligible.

Pratik Mazumder, Pravendra Singh, Vinay P. Namboodiri

In this paper, we propose an approach to improve few-shot classification performance using a composite rotation based auxiliary task. Few-shot classification methods aim to produce neural networks that perform well for classes with a large number of training samples and classes with less number of training samples. They employ techniques to enable the network to produce highly discriminative features that are also very generic. Generally, the better the quality and generic-nature of the features produced by the network, the better is the performance of the network on few-shot learning. Our approach aims to train networks to produce such features by using a self-supervised auxiliary task. Our proposed composite rotation based auxiliary task performs rotation at two levels, i.e., rotation of patches inside the image (inner rotation) and rotation of the whole image (outer rotation) and assigns one out of 16 rotation classes to the modified image. We then simultaneously train for the composite rotation prediction task along with the original classification task, which forces the network to learn high-quality generic features that help improve the few-shot classification performance. We experimentally show that our approach performs better than existing few-shot learning methods on multiple benchmark datasets.

Pravendra Singh, Pratik Mazumder, Vinay P. Namboodiri

This work presents a novel training technique for deep neural networks that makes use of additional data from a distribution that is different from that of the original input data. This technique aims to reduce overfitting and improve the generalization performance of the network. Our proposed technique, namely Passive Batch Injection Training Technique (PBITT), even reduces the level of overfitting in networks that already use the standard techniques for reducing overfitting such as $L_2$ regularization and batch normalization, resulting in significant accuracy improvements. Passive Batch Injection Training Technique (PBITT) introduces a few passive mini-batches into the training process that contain data from a distribution that is different from the input data distribution. This technique does not increase the number of parameters in the final model and also does not increase the inference (test) time but still improves the performance of deep CNNs. To the best of our knowledge, this is the first work that makes use of different data distribution to aid the training of convolutional neural networks (CNNs). We thoroughly evaluate the proposed approach on standard architectures: VGG, ResNet, and WideResNet, and on several popular datasets: CIFAR-10, CIFAR-100, SVHN, and ImageNet. We observe consistent accuracy improvement by using the proposed technique. We also show experimentally that the model trained by our technique generalizes well to other tasks such as object detection on the MS-COCO dataset using Faster R-CNN. We present extensive ablations to validate the proposed approach. Our approach improves the accuracy of VGG-16 by a significant margin of 2.1% over the CIFAR-100 dataset.

Pratik Mazumder, Pravendra Singh, Kranti Kumar Parida et al.

In this paper, we propose a novel approach for generalized zero-shot learning in a multi-modal setting, where we have novel classes of audio/video during testing that are not seen during training. We use the semantic relatedness of text embeddings as a means for zero-shot learning by aligning audio and video embeddings with the corresponding class label text feature space. Our approach uses a cross-modal decoder and a composite triplet loss. The cross-modal decoder enforces a constraint that the class label text features can be reconstructed from the audio and video embeddings of data points. This helps the audio and video embeddings to move closer to the class label text embedding. The composite triplet loss makes use of the audio, video, and text embeddings. It helps bring the embeddings from the same class closer and push away the embeddings from different classes in a multi-modal setting. This helps the network to perform better on the multi-modal zero-shot learning task. Importantly, our multi-modal zero-shot learning approach works even if a modality is missing at test time. We test our approach on the generalized zero-shot classification and retrieval tasks and show that our approach outperforms other models in the presence of a single modality as well as in the presence of multiple modalities. We validate our approach by comparing it with previous approaches and using various ablations.

Rajat, Munender Varshney, Pravendra Singh et al.

Multiple categories of objects are present in most images. Treating this as a multi-class classification is not justified. We treat this as a multi-label classification problem. In this paper, we further aim to minimize the supervision required for providing supervision in multi-label classification. Specifically, we investigate an effective class of approaches that associate a weak localization with each category either in terms of the bounding box or segmentation mask. Doing so improves the accuracy of multi-label categorization. The approach we adopt is one of active learning, i.e., incrementally selecting a set of samples that need supervision based on the current model, obtaining supervision for these samples, retraining the model with the additional set of supervised samples and proceeding again to select the next set of samples. A crucial concern is the choice of the set of samples. In doing so, we provide a novel insight, and no specific measure succeeds in obtaining a consistently improved selection criterion. We, therefore, provide a selection criterion that consistently improves the overall baseline criterion by choosing the top k set of samples for a varied set of criteria. Using this criterion, we are able to show that we can retain more than 98% of the fully supervised performance with just 20% of samples (and more than 96% using 10%) of the dataset on PASCAL VOC 2007 and 2012. Also, our proposed approach consistently outperforms all other baseline metrics for all benchmark datasets and model combinations.

Badri N. Patro, Mayank Lunayach, Vinay P. Namboodiri

Understanding and explaining deep learning models is an imperative task. Towards this, we propose a method that obtains gradient-based certainty estimates that also provide visual attention maps. Particularly, we solve for visual question answering task. We incorporate modern probabilistic deep learning methods that we further improve by using the gradients for these estimates. These have two-fold benefits: a) improvement in obtaining the certainty estimates that correlate better with misclassified samples and b) improved attention maps that provide state-of-the-art results in terms of correlation with human attention regions. The improved attention maps result in consistent improvement for various methods for visual question answering. Therefore, the proposed technique can be thought of as a recipe for obtaining improved certainty estimates and explanations for deep learning models. We provide detailed empirical analysis for the visual question answering task on all standard benchmarks and comparison with state of the art methods.

Badri N. Patro, Vinod K. Kurmi, Sandeep Kumar et al.

Generating natural questions from an image is a semantic task that requires using vision and language modalities to learn multimodal representations. Images can have multiple visual and language cues such as places, captions, and tags. In this paper, we propose a principled deep Bayesian learning framework that combines these cues to produce natural questions. We observe that with the addition of more cues and by minimizing uncertainty in the among cues, the Bayesian network becomes more confident. We propose a Minimizing Uncertainty of Mixture of Cues (MUMC), that minimizes uncertainty present in a mixture of cues experts for generating probabilistic questions. This is a Bayesian framework and the results show a remarkable similarity to natural questions as validated by a human study. We observe that with the addition of more cues and by minimizing uncertainty among the cues, the Bayesian framework becomes more confident. Ablation studies of our model indicate that a subset of cues is inferior at this task and hence the principled fusion of cues is preferred. Further, we observe that the proposed approach substantially improves over state-of-the-art benchmarks on the quantitative metrics (BLEU-n, METEOR, ROUGE, and CIDEr). Here we provide project link for Deep Bayesian VQG \url{https://delta-lab-iitk.github.io/BVQG/}

Vinay Kumar Verma, Pravendra Singh, Vinay P. Namboodiri et al.

We present a filter pruning approach for deep model compression, using a multitask network. Our approach is based on learning a a pruner network to prune a pre-trained target network. The pruner is essentially a multitask deep neural network with binary outputs that help identify the filters from each layer of the original network that do not have any significant contribution to the model and can therefore be pruned. The pruner network has the same architecture as the original network except that it has a multitask/multi-output last layer containing binary-valued outputs (one per filter), which indicate which filters have to be pruned. The pruner's goal is to minimize the number of filters from the original network by assigning zero weights to the corresponding output feature-maps. In contrast to most of the existing methods, instead of relying on iterative pruning, our approach can prune the network (original network) in one go and, moreover, does not require specifying the degree of pruning for each layer (and can learn it instead). The compressed model produced by our approach is generic and does not need any special hardware/software support. Moreover, augmenting with other methods such as knowledge distillation, quantization, and connection pruning can increase the degree of compression for the proposed approach. We show the efficacy of our proposed approach for classification and object detection tasks.

Pravendra Singh, Munender Varshney, Vinay P. Namboodiri

Researchers have proposed various activation functions. These activation functions help the deep network to learn non-linear behavior with a significant effect on training dynamics and task performance. The performance of these activations also depends on the initial state of the weight parameters, i.e., different initial state leads to a difference in the performance of a network. In this paper, we have proposed a cooperative initialization for training the deep network using ReLU activation function to improve the network performance. Our approach uses multiple activation functions in the initial few epochs for the update of all sets of weight parameters while training the network. These activation functions cooperate to overcome their drawbacks in the update of weight parameters, which in effect learn better "feature representation" and boost the network performance later. Cooperative initialization based training also helps in reducing the overfitting problem and does not increase the number of parameters, inference (test) time in the final model while improving the performance. Experiments show that our approach outperforms various baselines and, at the same time, performs well over various tasks such as classification and detection. The Top-1 classification accuracy of the model trained using our approach improves by 2.8% for VGG-16 and 2.1% for ResNet-56 on CIFAR-100 dataset.

Badri N. Patro, Dev Chauhan, Vinod K. Kurmi et al.

In this paper, we propose a method for obtaining sentence-level embeddings. While the problem of securing word-level embeddings is very well studied, we propose a novel method for obtaining sentence-level embeddings. This is obtained by a simple method in the context of solving the paraphrase generation task. If we use a sequential encoder-decoder model for generating paraphrase, we would like the generated paraphrase to be semantically close to the original sentence. One way to ensure this is by adding constraints for true paraphrase embeddings to be close and unrelated paraphrase candidate sentence embeddings to be far. This is ensured by using a sequential pair-wise discriminator that shares weights with the encoder that is trained with a suitable loss function. Our loss function penalizes paraphrase sentence embedding distances from being too large. This loss is used in combination with a sequential encoder-decoder network. We also validated our method by evaluating the obtained embeddings for a sentiment analysis task. The proposed method results in semantic embeddings and outperforms the state-of-the-art on the paraphrase generation and sentiment analysis task on standard datasets. These results are also shown to be statistically significant.

Badri N. Patro, Vinay P. Namboodiri

In this paper, we consider the problem of solving semantic tasks such as `Visual Question Answering' (VQA), where one aims to answers related to an image and `Visual Question Generation' (VQG), where one aims to generate a natural question pertaining to an image. Solutions for VQA and VQG tasks have been proposed using variants of encoder-decoder deep learning based frameworks that have shown impressive performance. Humans however often show generalization by relying on exemplar based approaches. For instance, the work by Tversky and Kahneman suggests that humans use exemplars when making categorizations and decisions. In this work, we propose the incorporation of exemplar based approaches towards solving these problems. Specifically, we incorporate exemplar based approaches and show that an exemplar based module can be incorporated in almost any of the deep learning architectures proposed in the literature and the addition of such a block results in improved performance for solving these tasks. Thus, just as the incorporation of attention is now considered de facto useful for solving these tasks, similarly, incorporating exemplars also can be considered to improve any proposed architecture for solving this task. We provide extensive empirical analysis for the same through various architectures, ablations, and state of the art comparisons.

Yatin Dandi, Aniket Das, Soumye Singhal et al.

In this work, we propose a modeling technique for jointly training image and video generation models by simultaneously learning to map latent variables with a fixed prior onto real images and interpolate over images to generate videos. The proposed approach models the variations in representations using residual vectors encoding the change at each time step over a summary vector for the entire video. We utilize the technique to jointly train an image generation model with a fixed prior along with a video generation model lacking constraints such as disentanglement. The joint training enables the image generator to exploit temporal information while the video generation model learns to flexibly share information across frames. Moreover, experimental results verify our approach's compatibility with pre-training on videos or images and training on datasets containing a mixture of both. A comprehensive set of quantitative and qualitative evaluations reveal the improvements in sample quality and diversity over both video generation and image generation baselines. We further demonstrate the technique's capabilities of exploiting similarity in features across frames by applying it to a model based on decomposing the video into motion and content. The proposed model allows minor variations in content across frames while maintaining the temporal dependence through latent vectors encoding the pose or motion features.