Abstract

Text-to-speech (TTS) systems are modelled as mel-synthesizers followed by speech-vocoders since the era of statistical TTS that is carried forward into neural designs. We propose an alternative approach to TTS modelling referred to as ParrotTTS borrowing from self-supervised learning (SSL) methods. ParrotTTS takes a two-step approach by initially training a speech-to-speech model on unlabelled data that is abundantly available, followed by a text-to-embedding model that leverages speech with aligned transcriptions to extend it to TTS. ParrotTTS achieves competitive mean opinion scores on naturalness compared to traditional TTS models but significantly improves over the latter's data efficiency of transcribed pairs and speaker adaptation without transcriptions. This further paves the path to training TTS models on generically trained SSL speech models.

Abstract

We focus on the problem of LiDAR point cloud based loop detection (or Finding) and closure (LDC) for mobile robots. State-of-the-art (SOTA) methods directly generate learned embeddings from a given point cloud, require large data augmentation, and are not robust to wide viewpoint variations in 6 Degrees-of-Freedom (DOF). Moreover, the absence of strong priors in an unstructured point cloud leads to highly inaccurate LDC. In this original approach, we propose independent roll and pitch canonicalization of point clouds using a common dominant ground plane. We discretize the canonicalized point clouds along the axis perpendicular to the ground plane leads to images similar to digital elevation maps (DEMs), which expose strong spatial priors in the scene. Our experiments show that LDC based on learnt embeddings from such DEMs is not only data efficient but also significantly more robust, and generalizable than the current SOTA. We report an (average precision for loop detection, mean absolute translation/rotation error) improvement of (8.4, 16.7/5.43)% on the KITTI08 sequence, and (11.0, 34.0/25.4)% on GPR10 sequence, over the current SOTA. To further test the robustness of our technique on point clouds in 6-DOF motion we create and opensource a custom dataset called LidarUrbanFly Dataset (LUF) which consists of point clouds obtained from a LiDAR mounted on a quadrotor. More details on our website https://gsc2001.github.io/FinderNet/

Abstract

The current study aims to understand whether the constraints in the well-defined problem solving task impacts the ill- defined problem solving task of a different knowledge do- main. We chose variants of Raven’s advanced progressive matrices(APM) for well-defined problem-solving. For the ill- defined problem solving task we have chosen creative reason- ing tasks (CRT) which is of same knowledge domain and al- ternate uses task(Guilford, 1967) of a different knowledge do- main. We have collected a sample of 44 participants. We ob- served a significant effect of the variant of APM on CRT per- formance. Higher number of rules in the CRT was observed when it was preceded by cAPM (Mdn = 2) compared to the classic APM (Mdn = 1). Further, we observed a higher CRT Relationship Score under cAPM (Mdn = 78.5) compared to the classic APM (Mdn = 50.5) puzzle conditions. However, the variant of APM did not show a significant effect on AUT, fluency score(p = 0.8 for independent t-test) and elaboration score (p = 0.48 for Mann-Whitney U test). The current results indicate that creativity is more domain-specific than domain- general.

Abstract

Learning embeddings of any data largely depends on the ability of the target space to capture semantic relations. The widely used Euclidean space, where embeddings are represented as point vectors, is known to be lacking in its potential to exploit complex structures and relations. Contrary to standard Euclidean embeddings, in this work, we embed point clouds as discrete probability distributions in Wasserstein space. We build a contrastive learning setup to learn Wasserstein embeddings that can be used as a pre-training method with or without supervision towards any downstream task. We show that the features captured by Wasserstein embeddings are better in preserving the point cloud geometry, including both global and local information, thus resulting in improved quality embeddings. We perform exhaustive experiments and demonstrate the effectiveness of our method for point cloud classification, transfer learning, segmentation, and interpolation tasks over multiple datasets including synthetic and realworld objects. We also compare against recent methods that use Wasserstein space and show that our method outperforms them in all downstream tasks. Additionally, our study reveals a promising interpretation of capturing critical points of point clouds that makes our proposed method self-explainable.

Abstract

Recently, the fundamental problem of unsupervised domain adaptation (UDA) on 3D point clouds has been motivated by a wide variety of applications in robotics, virtual reality, and scene understanding, to name a few. The point cloud data acquisition procedures manifest themselves as significant domain discrepancies and geometric variations among both similar and dissimilar classes. The standard domain adaptation methods developed for images do not directly translate to point cloud data because of their complex geometric nature. To address this challenge, we leverage the idea of multimodality and alignment between distributions. We propose a new UDA architecture for point cloud classification that benefits from multimodal contrastive learning to get better class separation in both domains individually. Further, the use of optimal transport (OT) aims at learning source and target data distributions jointly to reduce the cross-domain shift and provide a better alignment. We conduct a comprehensive empirical study on PointDA-10 and GraspNetPC-10 and show that our method achieves state-of-the-art performance on GraspNetPC-10 (with ≈ 4-12% margin) and best average performance on PointDA-10. Our ablation studies and decision boundary analysis also validate the significance of our contrastive learning module and OT alig

Abstract

Masonry is a traditional, extremely durable type of construction material, and many heritage masonry structures from different historical eras have endured adverse environmental conditions to varying degrees despite sometimes significantly losing their integrity, strength, and durability. Throughout history, the choice of masonry materials has largely been determined by availability, local geological formations, and environmental conditions. Although masonry has evolved over the centuries, the fundamentals of using stone, aggregate, or brick in combination with a binding material have not changed. It is not surprising that strong masonry materials were used to build some of the most notable historical buildings in human history that are still standing today. Resources nowadays are devoted to heritage architecture restoration and conservation because of its cultural significance. For the proper selection of construction techniques that could be applied to structural restoration or before taking any remedial action, a thorough understanding of masonry materials is required. Architectural surveys, the use of written evidence, laboratory investigations, and on-site tests can all be used to achieve this. Hence, the current study provides a thorough overview of the materials and their mechanical properties that have been developed over more than five thousand years for the construction of different structural components of masonry buildings that helped perform well during past seismic events.

Abstract

In the whole procedure of preservation, conservation, and restoration of monuments and historical buildings, structural restoration is considered an undesirable parameter, since it implies interventions on a large scale that might harm the authenticity of a culturally protected building. However, this type of intervention is an inevitable action, because it pertains to the safety of the building and, most importantly, to the safety of the users. In this respect, authenticity and safety are two concepts contradicting each other, and they are ensured by professionals of different origins and philosophies, namely archaeology & architecture on one side and science & technology on the other. In the case of structural restoration, these professionals are required to find a common space of co-existence so that the results of restoration satisfy both concepts, authenticity and safety. The present study gives an overview of thorough knowledge of construction techniques of masonry elements or systems or forms used in the past. In fact, in a period of more than five millennia, various techniques of construction have been developed as a result of big revolutions that have taken place in the history of human civilization. A deep knowledge of all the above is necessary for the suitable selection of present-day techniques and materials that might be used in structural restoration, where principles of reversible or irreversible techniques play a vital role together with the compatibility and durability of intervention materials. The main objective of this study is to guide the structural engineers involved in the structural repair of ancient monuments and buildings of national importance.

Abstract

The convergence of computer vision and advertising has sparked substantial interest lately. Existing advertisement datasets are either subsets of existing datasets with specialized annotations or feature diverse annotations without a cohesive taxonomy among ad images. Notably, no datasets encompass diverse advertisement styles or semantic grouping at various levels of granularity. Our work addresses this gap by introducing MAdVerse, an extensive, multilingual compilation of more than 50,000 ads from the web, social media websites, and e-newspapers. Advertisements are hierarchically grouped with uniform granularity into 11 categories, divided into 51 sub-categories, and 524 fine-grained brands at leaf level, each featuring ads in various languages. We provide comprehensive baseline classification results for prediction tasks within the realm of advertising analysis. These tasks include hierarchical ad classification, source classification, multilingual classification, and inducing hierarchy in existing ad datasets.

Abstract

We study the problem of fair sequential decision making given voter preferences. In each round, a decision rule must choose a decision from a set of alternatives where each voter reports which of these alternatives they approve. Instead of going with the most popular choice in each round, we aim for proportional representation, using axioms inspired by the multi-winner voting literature. The axioms require that every group of α% of the voters, if it agrees in every round (i.e., approves a common alternative), then those voters must approve at least α% of the decisions. A stronger version of the axioms requires that every group of α% of the voters that agrees in a β fraction of rounds must approve β · α% of the decisions. We show that three attractive voting rules satisfy axioms of this style. One of them (Sequential Phragmén) makes its decisions online, and the other two satisfy strengthened versions of the axioms but make decisions semi-online (Method of Equal Shares) or fully offline (Proportional Approval Voting). We present empirical results for these rules based on synthetic data and U.S. political elections. We also run experiments using the moral machine dataset about ethical dilemmas. We train preference models on user responses from different countries and let the models cast votes. We find that aggregating these votes using our rules leads to a more equal utility distribution across demographics than making decisions using a single global preference model.

Abstract

Radiance Fields are being widely explored for 3D scene reconstruction and several downstream tasks, such as segmentation. Prior radiance field segmentation methods require scene-specific training to enable segmentation. We propose distilling semantic features into Radiance Fields in a generalisable fashion using GNT, a transformer-based architecture, enabling 3D reconstruction and multi-view segmentation on arbitrarily new scenes. By fine-tuning our method, any set of 2D features can be distilled into a radiance field, providing better multi-view consistency than the original features. We show multi-view segmentation results on standard datasets and compare our method against existing NeRF-based segmentation methods. We perform on par with the state-of-the-art scene-specific segmentation methods. Our approach and experiments bring generalisable NeRF methods one step closer to the contemporary NeRF literature.