检索结果-内蒙古大学图书馆

27th International Conference on Pattern Recognition, ICPR 2024

作者： Dev, Prabhu Prasad Hazari, Raju Das, Pranesh Machine Learning Laboratory Department of Computer Science and Engineering National Institute of Technology Calicut India

ISBN: (纸本)9783031781247

Towards Video Anomaly Detection (VAD), existing methods require labor-intensive data collection and model retraining, making them costly and domain-specific. The proposed method, termed as Multi-modal Caption Aware Network (MCANet), introduces a novel paradigm that identifies anomalies in video sequences without requiring prior domain knowledge. This training-free VAD approach dynamically generates and analyzes textual descriptions of video frames by utilizing off-the-shelf vision-language model (VLM), audio-language model (ALM) and large language model (LLM). MCANet has four primary modules. The first module utilizes image-text similarities to clean noisy captions generated by the image captioning model, while the second module applies audio-text similarities to refine noisy captions produced by the audio captioning model. The third module employs a LLM to consolidate scene dynamics over time. Finally, the fourth module enhances the results by aggregating scores from semantically similar frames based on video-text similarity. To validate the effectiveness of the proposed method, experiments are conducted on two large-scale benchmark datasets (UCF-Crime and XD-Violence). Experimental results demonstrate that MCANet surpasses existing unsupervised and one-class approaches without requiring any training or data collection. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Hybrid speech enhancement in modulation domain

引用

Multimedia Tools and Applications 2024年 1-31页

作者： P S, Praveen Kumar H S, Jayanna Technical Lead-Machine Learning Merit Data and Technology Tamil Nadu Chennai India Department of Information Science and Engineering Siddaganga Institute of Technology Karnataka Tumkur India

This paper presents a comprehensive study on speech enhancement (SE) techniques, particularly focusing on the utilization of the discrete cosine transform (DCT) in the modulation domain (MD) in combination with the minimum mean square error (MMSE) and Kalman filter (KF) algorithms. The research investigates the efficacy of these techniques in improving speech quality and intelligibility under various noisy conditions, such as babble, drilling, horn, train, and traffic noises, across different signal-to-noise ratio (SNR) levels. Additionally, the study evaluates the performance of recurrent neural network (RNN) algorithms alongside traditional SE approaches. The study highlights the advantages of employing DCT in the modulation domain over the traditional discrete Fourier transform (DFT) in SE applications. Experimental results demonstrate significant improvements in objective quality metrics, including Perceptual Evaluation of Speech Quality (PESQ), composite measures, and short-time objective intelligibility (STOI), when implementing SE algorithms using discrete cosine transform in modulation domain. The proposed hybrid speech enhancement techniques, leveraging minimum mean square error and Kalman filter in combination with discrete cosine transform in modulation domain, outperform individual speech enhancement algorithms, showcasing superior noise reduction capabilities. This paper contributes to the advancement of speech enhancement methodologies, particularly in real-world noisy environments, and underscores the effectiveness of discrete cosine transform-based approaches in enhancing speech quality and intelligibility. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

An Empirical Analysis on Spatial Reasoning Capabilities of Large Multimodal Models

An Empirical Analysis on Spatial Reasoning Capabilities of L...

引用

2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024

作者： Shiri, Fatemeh Guo, Xiao-Yu Far, Mona Golestan Yu, Xin Haffari, Gholamreza Li, Yuan-Fang Department of Data Science & AI Monash University Australia Australian Institute for Machine Learning University of Adelaide Australia School of Electrical Engineering and Computer Science University of Queensland Australia

ISBN: (纸本)9798891761643

Large Multimodal Models (LMMs) have achieved strong performance across a range of vision and language tasks. However, their spatial reasoning capabilities are under-investigated. In this paper, we construct a novel VQA dataset, Spatial-MM, to comprehensively study LMMs' spatial understanding and reasoning capabilities. Our analyses on object-relationship and multi-hop reasoning reveal several important findings. Firstly, bounding boxes and scene graphs, even synthetic ones, can significantly enhance LMMs' spatial reasoning. Secondly, LMMs struggle more with questions posed from the human perspective than the camera perspective about the image. Thirdly, chain of thought (CoT) prompting does not improve model performance on complex multi-hop questions involving spatial relations. Lastly, our perturbation analysis on GQA-spatial reveals that LMMs are much stronger at basic object detection than complex spatial reasoning. We believe our new benchmark dataset and in-depth analyses can spark further research on LMMs spatial reasoning. © 2024 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

LSTM based Soft-Sensor for Estimating Nitrate Concentration in Aquaponics pond 3

LSTM based Soft-Sensor for Estimating Nitrate Concentration ...

引用

3rd International Conference for Innovation in Technology, INOCON 2024

作者： Dharshan, A. Kumar, Purushottam Ravimaran, S. Srinivasulu Reddy, U. Saranathan College of Engineering Department of Artificial Intelligence and Data Science Trichy India National Institute of Technology Artificial Intelligence Machine Learning & Data Analytics Lab Trichy India National Institute of Technology Department of Computer Applications Machine Learning & Data Analytics Lab Trichy India

ISBN: (纸本)9798350381931

In the field of aquaponics, where fish and plants coexist in a symbiotic environment, closely monitoring nitrate levels in the water is crucial due to their profound impact on aquatic and plant well-being. Traditional nitrate measurement methods are often time-consuming and costly. Various approaches, including first principles, IoT-based sensors, and machine learning-based soft sensors, have been attempted to address this challenge. However, these efforts face challenges such as expensive sensors, infrequent data collection, multistage data processing using limited sensor types, and the need for regular maintenance like cleaning and calibration. Additionally, varied environmental conditions affect sensor suitability for different water environments, and even some machine learning-based soft sensors have proven inaccurate. In response, soft sensors, especially deep learning-based ones, have gained prominence in industrial applications for their adaptability and accuracy. These sensors provide real-time insights into complex processes without requiring expensive hardware. In this study, an innovative solution was introduced using Long Short-Term Memory (LSTM) technology, a neural network architecture in deep learning known for capturing complex temporal patterns. LSTM is well-suited for modeling and predicting nitrate concentration changes in aquaponics, trained with extensive data collected from various aquaponic ponds. Through rigorous evaluation, a remarkable MSE value of 0.00074 and an impressive R-squared score of 0.98 were achieved, holding potential for scaling up to commercial applications, benefiting aquaponics operations, supporting researchers, and enhancing sustainability and productivity in aquaponic systems. © 2024 IEEE.

关键词： Nitrates

来源：评论

学校读者我要写书评

暂无评论

Enhancing Automated Short Answer Grading with Prompt-Driven Augmentation and Prompt Adaptive Oversampling 27th

Enhancing Automated Short Answer Grading with Prompt-Driven...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Afeefa, P.P. Hazari, Raju Das, Pranesh Machine Learning Laboratory Department of Computer Science and Engineering National Institute of Technology Calicut Kozhikode India

ISBN: (纸本)9783031781186

Automated Short Answer Grading (ASAG) comes under automatic answer script evaluation where the answer length is limited from one phrase to one paragraph. The main task in ASAG is generating a good sentence embedding for both the student and the reference answers. The existing works on the embedding creation perform better when using different deep-learning techniques and language models. However, the deep-learning techniques’ performance mainly depends on the training set size and quality. Most of the publicly available datasets typically have a limited number of reference and student answer pairs. To automate the dataset expansion, text augmentation techniques can be used. Conventional methods like back-translation, synonym replacement, and random deletion may replace some important technical words with other non-relevant terms, resulting in a loss of contextual meaning. We propose a new augmentation strategy for the ASAG datasets using LLM (Large Language Model) prompting. The effect of the proposed strategy is analysed on sentence transformer fine-tuning. We experimented with four different sizes of augmented training sets to determine the impact of the size of augmented training data on fine-tuning the sentence transformer model. Results indicate that sentence transformer fine-tuned using a 50% prompt-driven augmented dataset generates better embeddings. After having good embeddings, the traditional classifiers can be used to classify the student answers to different scores. We introduce "Prompt Adaptive Oversampling (PAO)" to address the class imbalance issue during grade classification. The effectiveness of the proposed strategy is analysed on two different public datasets: SPRAG, and Mohler-ASAG. The proposed method performs better while training highly imbalanced datasets. The source code of this work is available here. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Automated Plant Disease Detection: CNN for Corn Maize, Tomato, and Potato 7th

Automated Plant Disease Detection: CNN for Corn Maize, Tomat...

引用

7th International Conference on Soft Computing: Theories and Applications, SoCTA 2023

作者： Angeline, R. Aruneshwaran, S. Department of Computer Science and Engineering with Specialization in Artificial Intelligence and Machine Learning SRM Institute of Science and Technology Ramapuram India

ISBN: (纸本)9789819720880

Plant ailments pose present a significant challenge to the worldwide food security and the agricultural sector. Swift and precise detection of these diseases is pivotal for effectively managing them and preventing crop yield reductions. Lately, advanced deep learning techniques, specifically Convolutional Neural Networks (CNNs), have exhibited encouraging outcomes across various tasks involving image recognition. This undertaking strives to create and execute a model founded on CNNs to prognosticate plant diseases through leaf images. The proposed strategy encompasses three main phases: compiling and preparing the data, developing the model architecture, and assessing performance. Initially, an extensive dataset of plant leaf images, encompassing leaves afflicted by diverse diseases, is assembled. The images undergo preprocessing to heighten quality and eliminate disturbances, ensuring a dependable model training process. Subsequently, a CNN structure is devised and trained to employ the dataset. The chosen CNN model adheres to a sequential design, where each layer possesses precisely one input and output. These layers are arranged sequentially to construct the entire network and incorporate multiple convolutional layers such as Conv2D, MaxPooling2D, Flatten, and Dense, enabling the learning of features from the input images. The findings underscore that the CNN-centered model for forecasting plant diseases attains remarkable training precision of 99.65%, accompanied by a testing precision of 99.44% and a validation precision of 98.61%, proficiently identifying prevalent ailments like common rust disease in corn plants, bacterial spot infection in tomato crops, and the early blight ailment in potato plants. In conclusion, the proposed CNN-driven prognostic model for plant diseases manifests encouraging outcomes in precisely recognizing these diseases from leaf images. The efficacious application of this model can assist farmers and agricultural specialists in inform

关键词： Fruits

来源：评论

学校读者我要写书评

暂无评论

WHOLE-SONG HIERARCHICAL GENERATION OF SYMBOLIC MUSIC USING CASCADED DIFFUSION MODELS 12

WHOLE-SONG HIERARCHICAL GENERATION OF SYMBOLIC MUSIC USING C...

引用

12th International Conference on learning Representations, ICLR 2024

作者： Wang, Ziyu Min, Lejun Xia, Gus Computer Science Department NYU Shanghai China Machine Learning Department MBZUAI United Arab Emirates

Recent deep music generation studies have put much emphasis on long-term generation with structures. However, we are yet to see high-quality, well-structured whole-song generation. In this paper, we make the first attempt to model a full music piece under the realization of compositional hierarchy. With a focus on symbolic representations of pop songs, we define a hierarchical language, in which each level of hierarchy focuses on the semantics and context dependency at a certain music scope. The high-level languages reveal whole-song form, phrase, and cadence, whereas the low-level languages focus on notes, chords, and their local patterns. A cascaded diffusion model is trained to model the hierarchical language, where each level is conditioned on its upper levels. Experiments and analysis show that our model is capable of generating full-piece music with recognizable global verse-chorus structure and cadences, and the music quality is higher than the baselines. Additionally, we show that the proposed model is controllable in a flexible way. By sampling from the interpretable hierarchical languages or adjusting pre-trained external representations, users can control the music flow via various features such as phrase harmonic structures, rhythmic patterns, and accompaniment texture. © 2024 12th International Conference on learning Representations, ICLR 2024. All rights reserved.

关键词： Diffusion

来源：评论

学校读者我要写书评

暂无评论

Optimizing Crop Yields with Multilayer Perceptrons: A data-Driven Approach 8

Optimizing Crop Yields with Multilayer Perceptrons: A Data-D...

引用

8th International Conference on Inventive Systems and Control, ICISC 202

作者： Nivetha, N. Usharani, S. Department of Computer Science and Engineering Villupuram India Department of Artificial Intelligence and Machine Learning Villupuram India

ISBN: (纸本)9798350386578

Precision agriculture has become a major change in crop farming. It utilises cutting-edge technologies to maximise field-level management. Precision agriculture has completely transformed crop production by leveraging the latest developments to maximize field-level management. Predicting crop yields with accuracy helps farmers reduce their environmental impact, increase productivity, and make well-informed decisions. Accurate and timely insights are frequently lacking in traditional agricultural yield prediction approaches. The study offers a deep learning method for precisely predicting agricultural yields. Accurate crop yield forecasts assist farmers in minimizing their negative environmental effects, boosting productivity, and making educated choices. However, there are many obstacles because conventional agricultural yield prediction methods frequently need more timely and precise insights. Despite their success, several challenges still exist. These include handling heterogeneous data, dealing with missing values, and the complexity of capturing non-linear relationships in the data. To determine whether decision trees or Multi-Layer Perceptrons (MLP) are ideal in crop yield prediction, these models are compared with each other. Multi-layer perceptrons (MLP) are prominent among these techniques. Even though the MLP model was more accurate, decision trees also are relevant to the prediction process. This means have the capability of understanding multi-layer intra-data intricacies through their structure whereas decision trees may overfit on noisy data or grow too deep hence leading to many splits also known as being bushy unless they are pruned to reduce this bushiness. The study suggests a novel method for predicting agricultural productivity using a machine learning model Decision Tree and Multi-Layer Perceptrons (MLP). A web interface is also created to enable smooth communication with the prediction model. Through the usage of this interface, farmers and agr

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Using Generative Models to Improve Fire Detection Efficiency 10

Using Generative Models to Improve Fire Detection Efficiency

引用

10th International Conference on Information Technology and Nanotechnology, ITNT 2024

作者： Andriyanov, Nikita Kim, Alexandr Fao, Xenin Data Analysis And Machine Learning Department Financial University Under The Government Of The Russian Federation Moscow Russia Data Science Department Huazhong University Of Science And Technology Huazhong China

ISBN: (纸本)9798350372939

The paper discusses generative artificial intelligence technologies used to improve the efficiency of fire detection in satellite images. Different detector architectures are proposed and compared in terms of accuracy. It is found that the use of generative models to expand the training sample is on average 1-2% more effective than the use of standard augmentations. The combined method of expanding the training image base can improve the accuracy by 4-5%. In addition, the right approach in generative prompts can also improve the accuracy metrics of the models. © 2024 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Navigating the Maze of Explainable AI: A Systematic Approach to Evaluating Methods and Metrics 38

Navigating the Maze of Explainable AI: A Systematic Approach...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Klein, Lukas Lüth, Carsten Schlegel, Udo Bungert, Till El-Assady, Mennatallah Jäger, Paul Interactive Machine Learning Group Germany ETH Zürich Department of Computer Science Switzerland Germany Heidelberg University Department of Computer Science Germany University of Konstanz Department of Computer Science Germany

Explainable AI (XAI) is a rapidly growing domain with a myriad of proposed methods as well as metrics aiming to evaluate their efficacy. However, current studies are often of limited scope, examining only a handful of XAI methods and ignoring underlying design parameters for performance, such as the model architecture or the nature of input data. Moreover, they often rely on one or a few metrics and neglect thorough validation, increasing the risk of selection bias and ignoring discrepancies among metrics. These shortcomings leave practitioners confused about which method to choose for their problem. In response, we introduce LATEC, a large-scale benchmark that critically evaluates 17 prominent XAI methods using 20 distinct metrics. We systematically incorporate vital design parameters like varied architectures and diverse input modalities, resulting in 7,560 examined combinations. Through LATEC, we showcase the high risk of conflicting metrics leading to unreliable rankings and consequently propose a more robust evaluation scheme. Further, we comprehensively evaluate various XAI methods to assist practitioners in selecting appropriate methods aligning with their needs. Curiously, the emerging top-performing method, Expected Gradients, is not examined in any relevant related study. LATEC reinforces its role in future XAI research by publicly releasing all 326k saliency maps and 378k metric scores as a (meta-)evaluation dataset. The benchmark is hosted at: https://***/IML-DKFZ/latec. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：