Contrastive learning has emerged as a powerful technique in audio-visual representation learning, leveraging the natural co-occurrence of audio and visual modalities in extensive web-scale video datasets to achieve significant advancements. However, conventional contrastive audio-visual learning methodologies often rely on aggregated representations derived through temporal aggregation, which neglects the intrinsic sequential nature of the data. This oversight raises concerns regarding the ability of standard approaches to capture and utilize fine-grained information within sequences, information that is vital for distinguishing between semantically similar yet distinct examples. In response to this limitation, we propose sequential contrastive audio-visual learning (SCAV), which contrasts examples based on their non-aggregated representation space using sequential distances. Retrieval experiments with the VGGSound and Music datasets demonstrate the effectiveness of SCAV, showing 2-3x relative improvements against traditional aggregation-based contrastive learning and other methods from the literature. We also show that models trained with SCAV exhibit a high degree of flexibility regarding the metric employed for retrieval, allowing them to operate on a spectrum of efficiency-accuracy trade-offs, potentially making them applicable in multiple scenarios, from small- to large-scale retrieval.

In particular, we focus on neural networks that can dynamically activate or de-activate parts of their computational graph conditionally on their input. Examples include the dynamic selection of, e.g., input tokens, layers (or sets of layers), and sub-modules inside each layer (e.g., channels in a convolutional filter). We first provide a general formalism to describe these techniques in an uniform way. Then, we introduce three notable implementations of these principles: mixture-of-experts (MoEs) networks, token selection mechanisms, and early-exit neural networks. The paper aims to provide a tutorial-like introduction to this growing field. To this end, we analyze the benefits of these modular designs in terms of efficiency, explainability, and transfer learning, with a focus on emerging applicative areas ranging from automated scientific discovery to semantic communication.

Large language models (LLMs) have convincing performance in a variety of downstream tasks. However, these systems are prone to generating undesirable outputs such as harmful and biased text. In order to remedy such generations, the development of guardrail (or detector) models has gained traction. Motivated by findings from developing a detector for social bias, we adopt the notion of a use-mention distinction - which we identified as the primary source of under-performance in the preliminary versions of our social bias detector. Armed with this information, we describe a fully extensible and reproducible synthetic data generation pipeline which leverages taxonomy-driven instructions to create targeted and labeled data. Using this pipeline, we generate over 300K unique contrastive samples and provide extensive experiments to systematically evaluate performance on a suite of open source datasets. We show that our method achieves competitive performance with a fraction of the cost in compute and offers insight into iteratively developing efficient and capable guardrail models. Warning: This paper contains examples of text which are toxic, biased, and potentially harmful.

Large vision-language models (VLMs) can assist visually impaired people by describing images from their daily lives. Current evaluation datasets may not reflect diverse cultural user backgrounds or the situational context of this use case. To address this problem, we create a survey to determine caption preferences and propose a culture-centric evaluation benchmark by filtering VizWiz, an existing dataset with images taken by people who are blind. We then evaluate several VLMs, investigating their reliability as visual assistants in a culturally diverse setting. While our results for state-of-the-art models are promising, we identify challenges such as hallucination and misalignment of automatic evaluation metrics with human judgment. We make our survey, data, code, and model outputs publicly available.

The Music Emotion Recognition (MER) field has seen steady developments in recent years, with contributions from feature engineering, machine learning, and deep learning. The landscape has also shifted from audio-centric systems to bimodal ensembles that combine audio and lyrics. However, a severe lack of public and sizeable bimodal databases has hampered the development and improvement of bimodal audio-lyrics systems. This article proposes three new audio, lyrics, and bimodal MER research datasets, collectively called MERGE, created using a semi-automatic approach. To comprehensively assess the proposed datasets and establish a baseline for benchmarking, we conducted several experiments for each modality, using feature engineering, machine learning, and deep learning methodologies. In addition, we propose and validate fixed train-validate-test splits. The obtained results confirm the viability of the proposed datasets, achieving the best overall result of 79.21% F1-score for bimodal classification using a deep neural network.

Existing text-to-video diffusion models rely solely on text-only encoders for their pretraining. This limitation stems from the absence of large-scale multimodal prompt video datasets, resulting in a lack of visual grounding and restricting their versatility and application in multimodal integration. To address this, we construct a large-scale multimodal prompt dataset by employing retrieval methods to pair in-context examples with the given text prompts and then utilize a two-stage training strategy to enable diverse video generation tasks within the same model. In the first stage, we propose a multimodal conditional video generation framework for pretraining on these augmented datasets, establishing a foundational model for grounded video generation. Secondly, we finetune the model from the first stage on three video generation tasks, incorporating multi-modal instructions. This process further refines the model's ability to handle diverse inputs and tasks, ensuring seamless integration of multi-modal information. After this two-stage train-ing process, VIMI demonstrates multimodal understanding capabilities, producing contextually rich and personalized videos grounded in the provided inputs, as shown in Figure 1. Compared to previous visual grounded video generation methods, VIMI can synthesize consistent and temporally coherent videos with large motion while retaining the semantic control. Lastly, VIMI also achieves state-of-the-art text-to-video generation results on UCF101 benchmark.

Wireless Capsule Endoscopy (WCE) is highly valued for its non-invasive and painless approach, though its effectiveness is compromised by uneven illumination from hardware constraints and complex internal dynamics, leading to overexposed or underexposed images. While researchers have discussed the challenges of low-light enhancement in WCE, the issue of correcting for different exposure levels remains underexplored. To tackle this, we introduce EndoUIC, a WCE unified illumination correction solution using an end-to-end promptable diffusion transformer (DiT) model. In our work, the illumination prompt module shall navigate the model to adapt to different exposure levels and perform targeted image enhancement, in which the Adaptive Prompt Integration (API) and Global Prompt Scanner (GPS) modules shall further boost the concurrent representation learning between the prompt parameters and features. Besides, the U-shaped restoration DiT model shall capture the long-range dependencies and contextual information for unified illumination restoration. Moreover, we present a novel Capsule-endoscopy Exposure Correction (CEC) dataset, including ground-truth and corrupted image pairs annotated by expert photographers. Extensive experiments against a variety of state-of-the-art (SOTA) methods on four datasets showcase the effectiveness of our proposed method and components in WCE illumination restoration, and the additional downstream experiments further demonstrate its utility for clinical diagnosis and surgical assistance.

The ability to generate artificial human movement patterns while meeting location and time constraints is an important problem in the security community, particularly as it enables the study of the analog problem of detecting such patterns while maintaining privacy. We frame this problem as an instance of abduction guided by a novel parsimony function represented as an aggregate truth value over an annotated logic program. This approach has the added benefit of affording explainability to an analyst user. By showing that any subset of such a program can provide a lower bound on this parsimony requirement, we are able to abduce movement trajectories efficiently through an informed (i.e., A*) search. We describe how our implementation was enhanced with the application of multiple techniques in order to be scaled and integrated with a cloud-based software stack that included bottom-up rule learning, geolocated knowledge graph retrieval/management, and interfaces with government systems for independently conducted government-run tests for which we provide results. We also report on our own experiments showing that we not only provide exact results but also scale to very large scenarios and provide realistic agent trajectories that can go undetected by machine learning anomaly detectors.

Graph Neural Networks (GNNs) have become fundamental in graph-structured deep learning. Key paradigms of modern GNNs include message passing, graph rewiring, and Graph Transformers. This paper introduces Graph-Rewiring Attention with Stochastic Structures (GRASS), a novel GNN architecture that combines the advantages of these three paradigms. GRASS rewires the input graph by superimposing a random regular graph, enhancing long-range information propagation while preserving structural features of the input graph. It also employs a unique additive attention mechanism tailored for graph-structured data, providing a graph inductive bias while remaining computationally efficient. Our empirical evaluations demonstrate that GRASS achieves state-of-the-art performance on multiple benchmark datasets, confirming its practical efficacy.

Large Language Models have found application in various mundane and repetitive tasks including Human Resource (HR) support. We worked with the domain experts of SAP SE to develop an HR support chatbot as an efficient and effective tool for addressing employee inquiries. We inserted a human-in-the-loop in various parts of the development cycles such as dataset collection, prompt optimization, and evaluation of generated output. By enhancing the LLM-driven chatbot's response quality and exploring alternative retrieval methods, we have created an efficient, scalable, and flexible tool for HR professionals to address employee inquiries effectively. Our experiments and evaluation conclude that GPT-4 outperforms other models and can overcome inconsistencies in data through internal reasoning capabilities. Additionally, through expert analysis, we infer that reference-free evaluation metrics such as G-Eval and Prometheus demonstrate reliability closely aligned with that of human evaluation.