Abstract

This paper describes our system (iREL) for Tweet intimacy analysis shared task of the SemEval 2023 workshop at ACL 2023. Our system achieved an overall Pearson’s r score of 0.5924 and ranked 10th on the overall leaderboard. For the unseen languages, we ranked third on the leaderboard and achieved a Pearson’s r score of 0.485. We used a single multilingual model for all languages, as discussed in this paper. We provide a detailed description of our pipeline along with multiple ablation experiments to further analyse each component of the pipeline. We demonstrate how translation-based augmentation, domainspecific features, and domain-adapted pretrained models improve the understanding of intimacy in tweets. The code can be found at https://github.com/bhavyajeet/Multilingualtweet-intimacy

Abstract

The presence of a new decay mode relaxes the current mass exclusion limits on vectorlike quarks considerably. We consider the case of a weak-singlet vectorlike B quark that can decay to a singlet scalar or pseudoscalar Φ. In an earlier paper [A. Bhardwaj et al., Roadmap to explore vectorlike quarks decaying to a new scalar or pseudoscalar, Phys. Rev. D, 106 (2022) 095014; arXiv:2203.13753], we mapped the possibilities to explore such setups at the LHC. We showed that it is possible for a B quark to decay into Φ and the Φ to dominantly decay to a pair of gluons or b quark(s) without fine-tuning the parameters. In this paper, we present a collider search strategy to look for the pair production of singlet B quarks. If both B’s decay into bΦ pairs, the final state is fully hadronic: BB → (bΦ)(bΦ) → (bgg)(bgg)/(bbb)(bbb), which is very challenging to probe. Therefore, we consider a simpler mixed decay specific to the singlet B case, BB → (bΦ)(tW) with a lepton in the final state, to achieve the best sensitivity at the highluminosity LHC. We use a deep neural network with a weighted cross-entropy loss to separate the signal from the huge SM background. Our analysis shows that large areas of the MB − MΦ parameter space are discoverable through this signature. We show how the discovery and exclusion regions scale with the branching ratio in the new decay mode. We also estimate the reach in the inclusive monolepton channel with the same network model.

Abstract

Autonomous driving requires accurate reasoning of the location of objects from raw sensor data. Recent end-to-end learning methods go from raw sensor data to a trajectory output via Bird’s Eye View(BEV) segmentation as an interpretable intermediate representation. Motion planning over cost maps generated via Birds Eye View (BEV) segmentation has emerged as a prominent approach in autonomous driving. However, the current approaches have two critical gaps. First, the optimization process is simplistic and involves just evaluating a fixed set of trajectories over the cost map. The trajectory samples are not adapted based on their associated cost values. Second, the existing cost maps do not account for the uncertainty in the cost maps that can arise due to noise in RGB images, BEV annotations etc. As a result, these approaches can struggle in challenging scenarios where there is abrupt cutin, stopping, overtaking, merging, etc from the neighbouring vehicles. In this paper, we propose UAP-BEV, a novel approach that models the noise in Spatio-Temporal BEV predictions to create an uncertainty-aware occupancy grid map. Using queries of the distance to the closest occupied cell, we obtain a sample estimate of the collision probability of the ego-vehicle. Subsequently, our approach uses gradient-free sampling-based optimization to compute low-cost trajectories over the cost

Abstract

We propose a second order gradient based method with ADAM and RMSprop for the training of generative adversarial networks. The proposed method is fastest to obtain similar accuracy when compared to prominent second order methods. Unlike state-of-the-art recent methods, it does not require solving a linear system, or it does not require additional mixed second derivative terms. We derive the fixed point iteration corresponding to proposed method, and show that the proposed method is convergent. The proposed method produces better or comparable inception scores, and comparable quality of images compared to other recently proposed state-of-the-art second order methods. Compared to first order methods such as ADAM, it produces significantly better inception scores. The proposed method is compared and validated on popular datasets such as FFHQ, LSUN, CIFAR10, MNIST, and Fashion MNIST for image generation tasks1 . Codes: https://github.com/misterpawan/acom

Abstract

In our work, we propose a novel yet simple approach to obtain an adaptive learning rate for gradient-based descent methods on classification tasks. Instead of the traditional approach of selecting adaptive learning rates via the decayed expectation of gradient-based terms, we use the angle between the current gradient and the new gradient: this new gradient is computed from the direction orthogonal to the current gradient, which further helps us in determining a better adaptive learning rate based on angle history, thereby, leading to relatively better accuracy compared to the existing state-of-the-art optimizers. On a wide variety of benchmark datasets with prominent image classification architectures such as ResNet, DenseNet, EfficientNet, and VGG, we find that our method leads to the highest accuracy in most of the datasets. Moreover, we prove that our method is convergent. Index Terms—Adam, Image Classification, Optimization, Angle, Gradient Descent, Neural Networks, ResNet

Abstract

Extreme multi-label (XML) classification refers to the task of supervised multi-label learning that involves a large number of labels. Hence, scalability of the classifier with increasing label dimension is an important consideration. In this paper, we develop a method called LightDXML which modifies the recently developed deep learning based XML framework by using label embeddings instead of feature embedding for negative sampling and iterating cyclically through three major phases: (1) proxy training of label embeddings (2) shortlisting of labels for negative sampling and (3) final classifier training using the negative samples. Consequently, LightDXML also removes the requirement of a re-ranker module, thereby, leading to further savings on time and memory requirements. The proposed method achieves the best of both worlds: while the training time, model size and prediction times are on par or better compared to the tree-based methods, it attains much better prediction accuracy that is on par with the deep learning based methods. Moreover, the proposed approach achieves the best tail-label prediction accuracy over most state-of-the-art XML methods on some of the large datasets1 . Code: https://github.com/misterpawan/ LightDXML

Abstract

Over the years, the framework of Linear combination of unitaries (LCU) has been extremely useful for designing a plethora of quantum algorithms. In this work, we explore whether this widely applicable paradigm can be implemented on quantum computers that will be available immediately after the current NISQ stage. To this end, we develop three variants of LCU and apply each, to quantum algorithms of practical interest. First, we develop a physically motivated, continuous-time analogue of LCU (“Analog LCU”). This technique, implementable on hybrid qubit-qumode systems, is simpler than its discrete-time counterpart. We use this method to develop analog quantum algorithms for ground state preparation and quantum linear systems. We also develop a randomized quantum algorithm to sample from functions of Hamiltonians applied to quantum states (“Single-Ancilla LCU”). This approach repeatedly samples from a short-depth quantum circuit and uses only a single ancilla qubit. We use this to estimate expectation values of observables in the ground states of a Hamiltonian, and in the solution of quantum linear systems. This method is suitable for early fault-tolerant quantum computers. Our third approach stems from the observation that for several applications, it suffices to replace LCU with randomized sampling of unitaries according to the distribution of the LCU coefficients (“Ancilla-free LCU”). This is particularly useful when one is interested in the projection of a quantum state implemented by an LCU procedure in some subspace of interest. We demonstrate that this technique applies to the spatial search problem and helps establish a relationship between discrete and continuous-time quantum walks with their classical counterparts. Our work demonstrates that generic quantum algorithmic paradigms, such as LCU, can potentially be implemented on intermediate-term quantum devices.

Abstract

Throwing motion is known for phenomenally fast rearrangement, sorting tasks, and placing the object outside the limited workspace with less effort. However, in the robotics domain, despite many simple yet versatile, mechanically intelligent grippers reported earlier, they fo- cus primarily on achieving robust grasping and dexterous manipulation. This article presents a novel design of a sin- gle actuator driven hybrid gripper with mechanically cou- pled rigid links and elastic gripping surface; this arrange- ment provides the dual function of versatile grasping and throwing manipulation. The gripper comprises a latching mechanism (LM) that drives two passive rigid fingers by elongating/releasing the coupled elastic strip. Elongating the gripping surface enables the gripper to adapt to ob- jects with different geometries, vary surface contact force characteristics, and store the energy in the form of elastic potential. A mechanism to discharge the stored potential energy gradually or instantaneously is essential when the intended task is to place the object free from impact or away from the limited reachable workspace. The proposed LM can swiftly shift from a quick release to a gradual release of the stored elastic potential for greater object’s accelera- tion during throwing and no acceleration while placing. By doing so, the object can be placed at the desired location even farther than the manipulator’s reachable workspace. We report the proposed gripper’s design details, develop- ment, and experimentally demonstrate the versatile grasp- ing, impact-free placing, and throwing capabilities. Index Terms—Gripper design, multipurpose gripper, non- prehensile manipulation, throwing.

Abstract

Unmanned Aerial Vehicle (UAV) can be employed in various applications, including traffic control, and delivery application. However, these services are susceptible to various security attacks because sensitive data are exchanged through an open channel. Thus, a secure authentication scheme is essential for UAV. UAV has limited storage resources and computing capabilities. To overcome these problems, cloud computing is considered as a promising solution. Cloud computing provides various properties such as storage availability and scalability. Unfortunately, the cloud server’s database can be a major target for an adversary because it is a centralized system. If an adversary intrudes on the cloud server’s database, he/she may attempt to intercept or learn the stored data. To mitigate these issues, we design a secure and lightweight authentication and key agreement scheme for cloud-assisted UAV using blockchain in flying ad-hoc networks, called LAKA-UAV. LAKA-UAV utilizes blockchain technology to ensure access control and data integrity using log transactions and the cloud server securely manages collected data from UAV. We prove the security of LAKA-UAV based on informal security analysis and formal security verification implementation. Based on testbed experiments using MIRACL, LAKA-UAV provides about 2.13 times more efficient performance on average compared with related schemes in terms of computation. We present the blockchain implementation using Hyperledger Sawtooth platform

Abstract

Increasing use of electronic healthcare (eHealth) services demands efficient and secure solutions. Such solutions need to ensure the prevention of unauthorised access to the patient data and provide faster response. Fog computing is a viable solution to provide faster responses in eHealth systems. Key distribution and authentication play a major role in providing security to patient data. Existing centralized architectures are susceptible to single-point-of-compromise, that is, the entire system is vulnerable when the centralized authority keys are unexpectedly revealed to an adversary. In this paper, we present a fog-based semi-centralised architecture for key distribution and authentication, in which the key distribution service is delegated to individual fog-servers. Thus, the fog-servers become responsible for key distribution to the users without the involvement of the centralised authority, which forms a paradigm of multiple client-server architecture. Thus, achieving centralized trust by designing a single sign-on authentication in such environments is a challenging problem. We design a single sign-on authentication protocol for semi-centralized architectures to achieve centralized trust by ensuring that the user keys are independent of the centralized authority’s keys. A rigorous security analysis under the random oracle model is performed to prove that the proposed protocol is secure against single-point-of-compromise. We also conduct extensive experiments to show the practical perspectives of the proposed scheme. The results show that the protocol is suitable for eHealth applications, including emergency services.