Entries by NEC Labs America

Chain-of-region: Visual Language Models Need Details for Diagram Analysis

/in /by

Visual Language Models (VLMs) like GPT-4V have broadened the scope of LLM applications, yet they face significant challenges in accurately processing visual details, particularly in scientific diagrams. This paper explores the necessity of meticulous visual detail collection and region decomposition for enhancing the performance of VLMs in scientific diagram analysis. We propose a novel approach that combines traditional computer vision techniques with VLMs to systematically decompose diagrams into discernible visual elements and aggregate essential metadata. Our method employs techniques in OpenCV library to identify and label regions, followed by a refinement process using shape detection and region merging algorithms, which are particularly suited to the structured nature of scientific diagrams. This strategy not only improves the granularity and accuracy of visual information processing but also extends the capabilities of VLMs beyond their current limitations. We validate our approach through a series of experiments that demonstrate enhanced performance in diagram analysis tasks, setting a new standard for integrating visual and language processing in a multimodal context.

Progressive Token Length Scaling in Transformer Encoders for Efficient Universal Segmentation

/in /by

A powerful architecture for universal segmentation relies on transformers that encode multi-scale image features and decode object queries into mask predictions. With efficiency being a high priority for scaling such models, we observed that the state-of-the-art method Mask2Former uses >50% of its compute only on the transformer encoder. This is due to the retention of a full-length token-level representation of all backbone feature scales at each encoder layer. With this observation, we propose a strategy termed PROgressive Token Length SCALing for Efficient transformer encoders (PRO-SCALE) that can be plugged-in to the Mask2Former segmentation architecture to significantly reduce the computational cost. The underlying principle of PRO-SCALE is: progressively scale the length of the tokens with the layers of the encoder. This allows PRO-SCALE to reduce computations by a large margin with minimal sacrifice in performance (?52% encoder and ? 27% overall GFLOPs reduction with no drop in performance on COCO dataset). Experiments conducted on public benchmarks demonstrates PRO-SCALE’s flexibility in architectural configurations, and exhibits potential for extension beyond the settings of segmentation tasks to encompass object detection. Code available here: https://github.com/abhishekaich27/proscale-pytorch

TSLA: Unified Time Series and Language Model

/in /by

Real-world time series data often require analysis or interpretation from domain experts. Some tasks, like time series question answering, involve both time series and natural language questions, posing challenges for single-modality language models to understand their interaction. To this end, we present TSLA (Time Series Language Model), a framework designed to enhance the language model with the understanding of time series data for multi-modality tasks. TSLA comprises three key components. (1) Time Series Tokenizer learns how to represent time series data into discrete tokens, making it more manageable for language models. (2) Joint (Pre-)Training on task-agnostic time series and text data integrates time series tokens and text tokens to model the interplay between time series and language concepts. (3) Multi-task Instruction Tuning fine-tunes the pretrained TSLA for various downstream tasks relevant to user interests. For evaluation, we applied TSLA to time series data from human motions on four tasks: time series captioning, time series question answering, text-based time series synthesis, and text-based time series continuation. The results demonstrate TSLA’s effectiveness in handling multiple time series analysis tasks, pointing the way for future research endeavors.

Graph Neural Networks, Explained: Our Role in the Future of AI

/in /by

NEC Laboratories America (NECLA) is advancing the frontier of Graph Neural Networks (GNNs), a transformative AI technology that processes complex, interconnected data. Through innovations like PTDNet for robust learning, novel frameworks for explainability, StrGNN for anomaly detection in dynamic graphs, and GERDQ for calibration with out-of-distribution nodes, NECLA is addressing critical challenges in GNN development. These breakthroughs have real-world implications in fields such as cybersecurity, bioinformatics, and recommendation systems, positioning NECLA as a leader in the evolution of graph-based AI.

Trainingless Adaptation of Pretrained Models for Environmental Sound Classification

/in /by

Deep neural network (DNN)-based models for environmental sound classification are not robust against a domain to which training data do not belong, that is, out-of-distribution or unseen data. To utilize pretrained models for the unseen domain, adaptation methods, such as finetuning and transfer learning, are used with rich computing resources, e.g., the graphical processing unit (GPU). However, it is becoming more difficult to keep up with research trends for those who have poor computing resources because state-of-the-art models are becoming computationally resource-intensive. In this paper, we propose a trainingless adaptation method for pretrained models for environmental sound classification. To introduce the trainingless adaptation method, we first propose an operation of recovering time–frequency-ish (TF-ish) structures in intermediate layers of DNN models. We then propose the trainingless frequency filtering method for domain adaptation, which is not a gradient-based optimization widely used. The experiments conducted using the ESC-50 dataset show that the proposed adaptation method improves the classification accuracy by 20.40 percentage points compared with the conventional method.

Text-guided Device-realistic Sound Generation for Fiber-based Sound Event Classification

/in /by

Recent advancements in unique acoustic sensing devices and large-scale audio recognition models have unlocked new possibilities for environmental sound monitoring and detection. However, applying pretrained models to non-conventional acoustic sensors results in performance degradation due to domain shifts, caused by differences in frequency response and noise characteristics from the original training data. In this study, we introduce a text-guided framework for generating new datasets to retrain models specifically for these non-conventional sensors efficiently. Our approach integrates text-conditional audio generative models with two additional steps: (1) selecting audio samples based on text input to match the desired sounds, and (2) applying domain transfer techniques using recorded impulse responses and background noise to simulate the characteristics of the sensors. We demonstrate this process by generating emulated signals for fiber-optic Distributed Acoustic Sensors (DAS), creating datasets similar to the recorded ESC-50 dataset. The generated signals are then used to train a classifier, which outperforms few-shot learning approaches in environmental sound classification.

CLAP-S: Support Set Based Adaptation for Downstream Fiber-optic Acoustic Recognition

/in /by

Contrastive Language-Audio Pretraining (CLAP) models have demonstrated unprecedented performance in various acoustic signal recognition tasks. Fiber-optic-based acoustic recognition is one of the most important downstream tasks and plays a significant role in environmental sensing. Adapting CLAP for fiber-optic acoustic recognition has become an active research area. As a non-conventional acoustic sensor, fiberoptic acoustic recognition presents a challenging, domain-specific, low-shot deployment environment with significant domain shifts due to unique frequency response and noise characteristics. To address these challenges, we propose a support-based adaptation method, CLAP-S, which linearly interpolates a CLAP Adapter with the Support Set, leveraging both implicit knowledge through fine-tuning and explicit knowledge retrieved from memory for cross-domain generalization. Experimental results show that our method delivers competitive performance on both laboratory recorded fiber-optic ESC-50 datasets and a real-world fiber optic gunshot-firework dataset. Our research also provides valuable insights for other downstream acoustic recognition tasks.

On Synthesizing Data for Context Attribution in Question Answering

/in /by

Question Answering (QA) accounts for a significant portion of LLM usage “in the wild”. However, LLMs sometimes produce false or misleading responses, also known as “hallucinations”. Therefore, grounding the generated answers in contextually provided information — i.e., providing evidence for the generated text — is paramount for LLMs’ trustworthiness. Providing this information is the task of context attribution. In this paper, we systematically study LLM-based approaches for this task, namely we investigate (i) zero-shot inference, (ii) LLM ensembling, and (iii) fine-tuning of small LMs on synthetic data generated by larger LLMs. Our key contribution is SynQA: a novel generative strategy for synthesizing context attribution data. Given selected context sentences, an LLM generates QA pairs that are supported by these sentences. This leverages LLMs’ natural strengths in text generation while ensuring clear attribution paths in the synthetic training data. We show that the attribution data synthesized via SynQA is highly effective for fine-tuning small LMs for context attribution in different QA tasks and domains. Finally, with a user study, we validate the usefulness of small LMs (fine-tuned on synthetic data from SynQA) in context attribution for QA.

LLM-based Distributed Code Generation and Cost-Efficient Execution in the Cloud

/in /by

The advancement of Generative Artificial Intelligence (AI), particularly Large Language Models (LLMs), is reshaping the software industry by automating code generation. Many LLM-driven distributed processing systems rely on serial code generation constrained by predefined libraries, limiting flexibility and adaptability. While some approaches enhance performance through parallel execution or optimize edge-cloud distributed processing for specific domains, they often overlook the cost implications of deployment, restricting scalability and economic feasibility across diverse cloud environments. This paper presents DiCE-C, a system that eliminates these constraints by starting directly from a natural language query. DiCE-C dynamically identifies available tools at runtime, programmatically refines LLM prompts, and employs a stepwise approach—first generating serial code and then transforming it into distributed code. This adaptive methodology enables efficient distributed execution without dependence on specific libraries. By leveraging high-level parallelism at the Application Programming Interface (API) level and managing API execution as services within a Kubernetes-based runtime, DiCE-C reduces idle GPU time and facilitates the use of smaller, cost-effective GPU instances. Experiments with a vision-based insurance application demonstrate that DiCE-C reduces cloud operational costs by up to 72% when using smaller GPUs (A6000 and A4000 GPU machines vs. A100 GPU machine) and by 32% when using identical GPUs (A100 GPU machines). This flexible and cost-efficient approach makes DiCE-C a scalable solution for deploying LLM-generated vision applications in cloud environments.