Search

The deep learning language of choice these days is Python; measured by factors such as available libraries and technical support, it is hard to beat. At the same time, software written in lower-level programming languages like C++ retain advantages in speed. We describe a Python interface to Marian NMT, a C++-based training and inference toolkit for sequence-to-sequence models, focusing on machine translation. This interface enables models trained with Marian to be connected to the rich, wide range of tools available in Python. A highlight of the interface is the ability to compute state-of-the-art COMET metrics from Python but using Marian's inference engine, with a speedup factor of up to 7.8$\times$ the existing implementations. We also briefly spotlight a number of other integrations, including Jupyter notebooks, connection with prebuilt models, and a web app interface provided with the package. PyMarian is available in PyPI via $\texttt{pip install pymarian}$.

Utilizing large pre-trained models for specific tasks has yielded impressive results. However, fully fine-tuning these increasingly large models is becoming prohibitively resource-intensive. This has led to a focus on more parameter-efficient transfer learning, primarily within the same modality. But this approach has limitations, particularly in video understanding where suitable pre-trained models are less common. Addressing this, our study introduces a novel cross-modality transfer learning approach from images to videos, which we call parameter-efficient image-to-video transfer learning. We present the Facial-Emotion Adapter (FE-Adapter), designed for efficient fine-tuning in video tasks. This adapter allows pre-trained image models, which traditionally lack temporal processing capabilities, to analyze dynamic video content efficiently. Notably, it uses about 15 times fewer parameters than previous methods, while improving accuracy. Our experiments in video emotion recognition demonstrate that the FE-Adapter can match or even surpass existing fine-tuning and video emotion models in both performance and efficiency. This breakthrough highlights the potential for cross-modality approaches in enhancing the capabilities of AI models, particularly in fields like video emotion analysis where the demand for efficiency and accuracy is constantly rising.

The Segment Anything Model (SAM) has achieved remarkable successes in the realm of natural image segmentation, but its deployment in the medical imaging sphere has encountered challenges. Specifically, the model struggles with medical images that feature low contrast, faint boundaries, intricate morphologies, and small-sized objects. To address these challenges and enhance SAM's performance in the medical domain, we introduce a comprehensive modification. Firstly, we incorporate a frozen Convolutional Neural Network (CNN) branch as an image encoder, which synergizes with SAM's original Vision Transformer (ViT) encoder through a novel variational attention fusion module. This integration bolsters the model's capability to capture local spatial information, which is often paramount in medical imagery. Moreover, to further optimize SAM for medical imaging, we introduce feature and position adapters within the ViT branch, refining the encoder's representations. We see that compared to current prompting strategies to fine-tune SAM for ultrasound medical segmentation, the use of text descriptions that serve as text prompts for SAM helps significantly improve the performance. Leveraging ChatGPT's natural language understanding capabilities, we generate prompts that offer contextual information and guidance to SAM, enabling it to better understand the nuances of ultrasound medical images and improve its segmentation accuracy. Our method, in its entirety, represents a significant stride towards making universal image segmentation models more adaptable and efficient in the medical domain., Comment: Accepted to ECCV 2024

This is the preliminary ranking of WMT24 General MT systems based on automatic metrics. The official ranking will be a human evaluation, which is superior to the automatic ranking and supersedes it. The purpose of this report is not to interpret any findings but only provide preliminary results to the participants of the General MT task that may be useful during the writing of the system submission.

Following the rapid progress in natural language processing (NLP) models, language models are applied to increasingly more complex interactive tasks such as negotiations and conversation moderations. Having human evaluators directly interact with these NLP models is essential for adequately evaluating the performance on such interactive tasks. We develop BotEval, an easily customizable, open-source, evaluation toolkit that focuses on enabling human-bot interactions as part of the evaluation process, as opposed to human evaluators making judgements for a static input. BotEval balances flexibility for customization and user-friendliness by providing templates for common use cases that span various degrees of complexity and built-in compatibility with popular crowdsourcing platforms. We showcase the numerous useful features of BotEval through a study that evaluates the performance of various chatbots on their effectiveness for conversational moderation and discuss how BotEval differs from other annotation tools., Comment: ACL 2024 SDT, 10 pages

Contemporary medical contrastive learning faces challenges from inconsistent semantics and sample pair morphology, leading to dispersed and converging semantic shifts. The variability in text reports, due to multiple authors, complicates semantic consistency. To tackle these issues, we propose a two-step approach. Initially, text reports are converted into a standardized triplet format, laying the groundwork for our novel concept of ``observations'' and ``verdicts''. This approach refines the {Entity, Position, Exist} triplet into binary questions, guiding towards a clear ``verdict''. We also innovate in visual pre-training with a Meijering-based masking, focusing on features representative of medical images' local context. By integrating this with our text conversion method, our model advances cross-modal representation in a multimodal contrastive learning framework, setting new benchmarks in medical image analysis., Comment: Accepted in MICCAI-24

This paper focuses on the educational journey of a computer engineering undergraduate student venturing into the domain of computer vision and robotics. It explores how optical flow and its applications can be used to detect moving objects when a camera undergoes translational motion, highlighting the challenges encountered and the strategies used to overcome them. Furthermore, the paper discusses not only the technical skills acquired by the student but also interpersonal skills as related to teamwork and diversity. In this paper, we detail the learning process, including the acquisition of technical and problem-solving skills, as well as out-of-the-box thinking., Comment: 20 Pages

A vast majority of electrical devices have integrated magnetic units, which generate constant magnetic fields with noticeable vibrations. The majority of existing nanogenerators acquire energy through friction/mechanical forces and most of these instances overlook acoustic vibrations and magnetic fields. Magnetic two-dimensional (2D) tellurides present a wide range of possibilities for devising a potential flexible energy harvester. We have synthesized two-dimensional chromium telluride (2D CrTe3) which exhibits ferromagnetic (FM) nature with a Tc of 224 K. The structure exhibits stable high remnant magnetization, making 2D CrTe3 flakes a potential material for harvesting of magneto-acoustic waves at room temperature. A magneto-acoustic nanogenerator (MANG) was fabricated composing of 2D CrTe3 dispersed in a polymer matrix. Basic mechanical stability and sensitivity of the device with change in load conditions were tested. A high surface charge density of 2.919 mC m-2 was obtained for the device. The thermal strain created in the lattice structure was examined using in-situ Raman spectroscopic measurements. The magnetic anisotropy energy (MAE) responsible for long-range FM ordering was calculated with the help of theoretical modelling. The theoretical calculations also showed opening of electronic bandgap which enhances the flexoelectric effects. The MANG can be a potential energy harvester to synergistically tap into the magneto-acoustic vibrations generated from the frequency changes of a vibrating device such as loudspeakers.

Two-dimensional (2D) chromium telluride Cr2Te3 exhibits strong ferromagnetic ordering with high coercivity at low temperatures and paramagnetic behaviour when approaching room temperature. The spin states of monolayer Cr2Te3 show ferromagnetic ordering in the ground state, and in-situ Raman analysis shows reversible structure transformation and hence a ferromagnetic to paramagnetic transition during low-temperature heating cycles (0 - 25 {\deg}C). The magnetic phase transition near room temperature in the 2D Cr2Te3 prompted the exploration of these layered materials for energy application. We demonstrate that the low-temperature ferromagnetic behavior can be used to magnetically deice material surfaces using an external magnetic source, avoiding the use of harsh chemicals and high temperatures. The hydrophobic nature and dipole interactions of H2O molecules with the surface of the 2D Cr2Te3 coating aid in the condensation of ice droplets formed on the surface. First-principles calculations also confirm the observed crystal structure, surface interaction, and magnetic properties of 2D Cr2Te3.

Audio classification models, particularly the Audio Spectrogram Transformer (AST), play a crucial role in efficient audio analysis. However, optimizing their efficiency without compromising accuracy remains a challenge. In this paper, we introduce FastAST, a framework that integrates Token Merging (ToMe) into the AST framework. FastAST enhances inference speed without requiring extensive retraining by merging similar tokens in audio spectrograms. Furthermore, during training, FastAST brings about significant speed improvements. The experiments indicate that FastAST can increase audio classification throughput with minimal impact on accuracy. To mitigate the accuracy impact, we integrate Cross-Model Knowledge Distillation (CMKD) into the FastAST framework. Integrating ToMe and CMKD into AST results in improved accuracy compared to AST while maintaining faster inference speeds. FastAST represents a step towards real-time, resource-efficient audio analysis., Comment: Accepted to Interspeech 2024

We give an efficient algorithm for Centroid-Linkage Hierarchical Agglomerative Clustering (HAC), which computes a $c$-approximate clustering in roughly $n^{1+O(1/c^2)}$ time. We obtain our result by combining a new Centroid-Linkage HAC algorithm with a novel fully dynamic data structure for nearest neighbor search which works under adaptive updates. We also evaluate our algorithm empirically. By leveraging a state-of-the-art nearest-neighbor search library, we obtain a fast and accurate Centroid-Linkage HAC algorithm. Compared to an existing state-of-the-art exact baseline, our implementation maintains the clustering quality while delivering up to a $36\times$ speedup due to performing fewer distance comparisons.

Zero-Shot Learning (ZSL) aims to enable classifiers to identify unseen classes by enhancing data efficiency at the class level. This is achieved by generating image features from pre-defined semantics of unseen classes. However, most current approaches heavily depend on the number of samples from seen classes, i.e. they do not consider instance-level effectiveness. In this paper, we demonstrate that limited seen examples generally result in deteriorated performance of generative models. To overcome these challenges, we propose ZeroDiff, a Diffusion-based Generative ZSL model. This unified framework incorporates diffusion models to improve data efficiency at both the class and instance levels. Specifically, for instance-level effectiveness, ZeroDiff utilizes a forward diffusion chain to transform limited data into an expanded set of noised data. For class-level effectiveness, we design a two-branch generation structure that consists of a Diffusion-based Feature Generator (DFG) and a Diffusion-based Representation Generator (DRG). DFG focuses on learning and sampling the distribution of cross-entropy-based features, whilst DRG learns the supervised contrastive-based representation to boost the zero-shot capabilities of DFG. Additionally, we employ three discriminators to evaluate generated features from various aspects and introduce a Wasserstein-distance-based mutual learning loss to transfer knowledge among discriminators, thereby enhancing guidance for generation. Demonstrated through extensive experiments on three popular ZSL benchmarks, our ZeroDiff not only achieves significant improvements over existing ZSL methods but also maintains robust performance even with scarce training data. Code will be released upon acceptance.

This paper offers a comprehensive analysis of recent advancements in video inpainting techniques, a critical subset of computer vision and artificial intelligence. As a process that restores or fills in missing or corrupted portions of video sequences with plausible content, video inpainting has evolved significantly with the advent of deep learning methodologies. Despite the plethora of existing methods and their swift development, the landscape remains complex, posing challenges to both novices and established researchers. Our study deconstructs major techniques, their underpinning theories, and their effective applications. Moreover, we conduct an exhaustive comparative study, centering on two often-overlooked dimensions: visual quality and computational efficiency. We adopt a human-centric approach to assess visual quality, enlisting a panel of annotators to evaluate the output of different video inpainting techniques. This provides a nuanced qualitative understanding that complements traditional quantitative metrics. Concurrently, we delve into the computational aspects, comparing inference times and memory demands across a standardized hardware setup. This analysis underscores the balance between quality and efficiency: a critical consideration for practical applications where resources may be constrained. By integrating human validation and computational resource comparison, this survey not only clarifies the present landscape of video inpainting techniques but also charts a course for future explorations in this vibrant and evolving field.

Computing a Single-Linkage Dendrogram (SLD) is a key step in the classic single-linkage hierarchical clustering algorithm. Given an input edge-weighted tree $T$, the SLD of $T$ is a binary dendrogram that summarizes the $n-1$ clusterings obtained by contracting the edges of $T$ in order of weight. Existing algorithms for computing the SLD all require $\Omega(n\log n)$ work where $n = |T|$. Furthermore, to the best of our knowledge no prior work provides a parallel algorithm obtaining non-trivial speedup for this problem. In this paper, we design faster parallel algorithms for computing SLDs both in theory and in practice based on new structural results about SLDs. In particular, we obtain a deterministic output-sensitive parallel algorithm based on parallel tree contraction that requires $O(n \log h)$ work and $O(\log^2 n \log^2 h)$ depth, where $h$ is the height of the output SLD. We also give a deterministic bottom-up algorithm for the problem inspired by the nearest-neighbor chain algorithm for hierarchical agglomerative clustering, and show that it achieves $O(n\log h)$ work and $O(h \log n)$ depth. Our results are based on a novel divide-and-conquer framework for building SLDs, inspired by divide-and-conquer algorithms for Cartesian trees. Our new algorithms can quickly compute the SLD on billion-scale trees, and obtain up to 150x speedup over the highly-efficient Union-Find algorithm typically used to compute SLDs in practice., Comment: To appear at SPAA 2024

Average linkage Hierarchical Agglomerative Clustering (HAC) is an extensively studied and applied method for hierarchical clustering. Recent applications to massive datasets have driven significant interest in near-linear-time and efficient parallel algorithms for average linkage HAC. We provide hardness results that rule out such algorithms. On the sequential side, we establish a runtime lower bound of $n^{3/2-\epsilon}$ on $n$ node graphs for sequential combinatorial algorithms under standard fine-grained complexity assumptions. This essentially matches the best-known running time for average linkage HAC. On the parallel side, we prove that average linkage HAC likely cannot be parallelized even on simple graphs by showing that it is CC-hard on trees of diameter $4$. On the possibility side, we demonstrate that average linkage HAC can be efficiently parallelized (i.e., it is in NC) on paths and can be solved in near-linear time when the height of the output cluster hierarchy is small., Comment: To appear at ICALP 2024

Self-supervised learning (SSL) has emerged as a promising solution for addressing the challenge of limited labeled data in deep neural networks (DNNs), offering scalability potential. However, the impact of design dependencies within the SSL framework remains insufficiently investigated. In this study, we comprehensively explore SSL behavior across a spectrum of augmentations, revealing their crucial role in shaping SSL model performance and learning mechanisms. Leveraging these insights, we propose a novel learning approach that integrates prior knowledge, with the aim of curtailing the need for extensive data augmentations and thereby amplifying the efficacy of learned representations. Notably, our findings underscore that SSL models imbued with prior knowledge exhibit reduced texture bias, diminished reliance on shortcuts and augmentations, and improved robustness against both natural and adversarial corruptions. These findings not only illuminate a new direction in SSL research, but also pave the way for enhancing DNN performance while concurrently alleviating the imperative for intensive data augmentation, thereby enhancing scalability and real-world problem-solving capabilities.

The CIRCLES project aims to reduce instabilities in traffic flow, which are naturally occurring phenomena due to human driving behavior. These "phantom jams" or "stop-and-go waves,"are a significant source of wasted energy. Toward this goal, the CIRCLES project designed a control system referred to as the MegaController by the CIRCLES team, that could be deployed in real traffic. Our field experiment leveraged a heterogeneous fleet of 100 longitudinally-controlled vehicles as Lagrangian traffic actuators, each of which ran a controller with the architecture described in this paper. The MegaController is a hierarchical control architecture, which consists of two main layers. The upper layer is called Speed Planner, and is a centralized optimal control algorithm. It assigns speed targets to the vehicles, conveyed through the LTE cellular network. The lower layer is a control layer, running on each vehicle. It performs local actuation by overriding the stock adaptive cruise controller, using the stock on-board sensors. The Speed Planner ingests live data feeds provided by third parties, as well as data from our own control vehicles, and uses both to perform the speed assignment. The architecture of the speed planner allows for modular use of standard control techniques, such as optimal control, model predictive control, kernel methods and others, including Deep RL, model predictive control and explicit controllers. Depending on the vehicle architecture, all onboard sensing data can be accessed by the local controllers, or only some. Control inputs vary across different automakers, with inputs ranging from torque or acceleration requests for some cars, and electronic selection of ACC set points in others. The proposed architecture allows for the combination of all possible settings proposed above. Most configurations were tested throughout the ramp up to the MegaVandertest.

We address causal reasoning in multivariate time series data generated by stochastic processes. Existing approaches are largely restricted to static settings, ignoring the continuity and emission of variations across time. In contrast, we propose a learning paradigm that directly establishes causation between events in the course of time. We present two key lemmas to compute causal contributions and frame them as reinforcement learning problems. Our approach offers formal and computational tools for uncovering and quantifying causal relationships in diffusion processes, subsuming various important settings such as discrete-time Markov decision processes. Finally, in fairly intricate experiments and through sheer learning, our framework reveals and quantifies causal links, which otherwise seem inexplicable., Comment: Accepted at the Twelfth International Conference on Learning Representations (ICLR'24)

Deep learning models have revolutionized various fields, from image recognition to natural language processing, by achieving unprecedented levels of accuracy. However, their increasing energy consumption has raised concerns about their environmental impact, disadvantaging smaller entities in research and exacerbating global energy consumption. In this paper, we explore the trade-off between model accuracy and electricity consumption, proposing a metric that penalizes large consumption of electricity. We conduct a comprehensive study on the electricity consumption of various deep learning models across different GPUs, presenting a detailed analysis of their accuracy-efficiency trade-offs. By evaluating accuracy per unit of electricity consumed, we demonstrate how smaller, more energy-efficient models can significantly expedite research while mitigating environmental concerns. Our results highlight the potential for a more sustainable approach to deep learning, emphasizing the importance of optimizing models for efficiency. This research also contributes to a more equitable research landscape, where smaller entities can compete effectively with larger counterparts. This advocates for the adoption of efficient deep learning practices to reduce electricity consumption, safeguarding the environment for future generations whilst also helping ensure a fairer competitive landscape., Comment: Accepted to ECCV-GreenFOMO Workshop