检索结果-内蒙古大学图书馆

Transfer Learning-Based Prediction of Heating Energy Consumption Using Urban Building Energy Models (UBEM) 31

学校读者我要写书评

暂无评论

Transfer Learning-Based Prediction of Heating Energy Consump...

31st International Workshop on Intelligent Computing in engineering, EG-ICE 2024

作者： Canli, Ilkim Ucar, Sevval Gursoy, Fatma Ece Halacli, Eren Gokberk Iseri, Orcun Koral Yavuz, Feyza Dino, Ipek Gursel Kalkan, Sinan Middle East Technical University Department of Architecture Turkey Middle East Technical University Department of Computer Engineering Turkey ODTU-GUNAM Solar Energy Research and Applications Turkey METU-ROMER METU Robotics and AI Technologies Application and Research Center Turkey

To estimate the energy demand of buildings, machine learning (ML) using datasets generated by Urban Building Energy Modeling (UBEM) has emerged as a potent approach. However, collecting large training datasets for each urban context is challenging, and developing large UBEMs for each city is too resource-intensive. Transfer learning (TL) can address these limitations by transferring knowledge from an existing model to train a model in target cities with limited datasets. In this paper, we present an approach for TL-based prediction of heating energy consumption on an urban scale based on three existing UBEMs of cities with different climatic conditions and thermal characteristics of buildings. Our experiments show that TL does provide better estimation compared to training ML models from scratch. Furthermore, our results suggest that this gain in accuracy is even higher if the target city has a substantially low number of buildings to train an ML model. © 2024 Proceedings of the 31st International Workshop on Intelligent Computing in engineering, EG-ICE 2024. All rights reserved.

关键词： Adversarial machine learning

MoCap-to-Visual Domain Adaptation for Efficient Human Mesh Estimation from 2D Keypoints

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Uguz, Bedirhan Suat, Ozhan Karagoz, Batuhan Akbas, Emre Department of Computer Engineering METU ROMER Robotics Center Middle East Technical University Ankara Turkey

This paper presents Key2Mesh, a model that takes a set of 2D human pose keypoints as input and estimates the corresponding body mesh. Since this process does not involve any visual (i.e. RGB image) data, the model can be trained on large-scale motion capture (MoCap) datasets, thereby overcoming the scarcity of image datasets with 3D labels. To enable the model’s application on RGB images, we first run an off-the-shelf 2D pose estimator to obtain the 2D keypoints, and then feed these 2D keypoints to Key2Mesh. To improve the performance of our model on RGB images, we apply an adversarial domain adaptation (DA) method to bridge the gap between the MoCap and visual domains. Crucially, our DA method does not require 3D labels for visual data, which enables adaptation to target sets without the need for costly labels. We evaluate Key2Mesh for the task of estimating 3D human meshes from 2D keypoints, in the absence of RGB and mesh label pairs. Our results on widely used H3.6M and 3DPW datasets show that Key2Mesh sets the new state-of-the-art by outperforming other models in PA-MPJPE for both datasets, and in MPJPE and PVE for the 3DPW dataset. Thanks to our model’s simple architecture, it operates at least 12× faster than the prior state-of-the-art model, LGD [44]. Additional qualitative samples and code are available on the project website: https://***/. © 2024, CC BY.

关键词： Image enhancement

RankED: Addressing Imbalance and Uncertainty in Edge Detection Using Ranking-based Losses

学校读者我要写书评

暂无评论

RankED: Addressing Imbalance and Uncertainty in Edge Detecti...

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Bedrettin Cetinkaya Sinan Kalkan Emre Akbas Department of Computer Engineering METU ROMER Robotics Center Middle East Technical University Ankara Turkey

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Detecting edges in images suffers from the problems of (PI) heavy imbalance between positive and negative classes as well as (P2) label uncertainty owing to disagreement be-tween different annotators. Existing solutions address PI using class-balanced cross-entropy loss and dice loss and P2 by only predicting edges agreed upon by most annota-tors. In this paper, we propose RankED, a unified ranking-based approach that addresses both the imbalance problem (P 1) and the uncertainty problem (P2). RankED tackles these two problems with two components: One component which ranks positive pixels over negative pixels, and the second which promotes high confidence edge pixels to have more label certainty. We show that RankED outperforms previous studies and sets a new state-of-the-art on NYUD-v2, BSDS500 and Multi-cue datasets. Code is available at https://***.

关键词： Training computer vision Uncertainty Codes Image edge detection Detectors Pattern recognition

RankED: Addressing Imbalance and Uncertainty in Edge Detection Using Ranking-based Losses

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Cetinkaya, Bedrettin Kalkan, Sinan Akbas, Emre Department of Computer Engineering and METU ROMER Robotics Center Middle East Technical University Ankara Turkey

Detecting edges in images suffers from the problems of (P1) heavy imbalance between positive and negative classes as well as (P2) label uncertainty owing to disagreement between different annotators. Existing solutions address P1 using class-balanced cross-entropy loss and dice loss and P2 by only predicting edges agreed upon by most annotators. In this paper, we propose RankED, a unified ranking-based approach that addresses both the imbalance problem (P1) and the uncertainty problem (P2). RankED tackles these two problems with two components: One component which ranks positive pixels over negative pixels, and the second which promotes high confidence edge pixels to have more label certainty. We show that RankED outperforms previous studies and sets a new state-of-the-art on NYUDv2, BSDS500 and Multi-cue datasets. Code is available at https://***. © 2024, CC BY.

关键词： Pixels

Bucketed Ranking-based Losses for Efficient Training of Object Detectors

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yavuz, Feyza Cam, Baris Can Dogan, Adnan Harun Oksuz, Kemal Akbas, Emre Kalkan, Sinan Department of Computer Engineering METU Ankara Turkey Five AI Ltd. United Kingdom METU ROMER Robotics Center Ankara Turkey

Ranking-based loss functions, such as Average Precision Loss and Rank&Sort Loss, outperform widely used score-based losses in object detection. These loss functions better align with the evaluation criteria, have fewer hyperparameters, and offer robustness against the imbalance between positive and negative classes. However, they require pairwise comparisons among P positive and N negative predictions, introducing a time complexity of O(PN), which is prohibitive since N is often large (e.g., 108 in ATSS). Despite their advantages, the widespread adoption of ranking-based losses has been hindered by their high time and space complexities. In this paper, we focus on improving the efficiency of ranking-based loss functions. To this end, we propose Bucketed Ranking-based Losses which group negative predictions into B buckets (B N) in order to reduce the number of pairwise comparisons so that time complexity can be reduced. Our method enhances the time complexity, reducing it to O(max(N log(N), P2)). To validate our method and show its generality, we conducted experiments on 2 different tasks, 3 different datasets, 7 different detectors. We show that Bucketed Ranking-based (BR) Losses yield the same accuracy with the unbucketed versions and provide 2× faster training on average. We also train, for the first time, transformer-based object detectors using ranking-based losses, thanks to the efficiency of our BR. When we train CoDETR, a state-of-the-art transformer-based object detector, using our BR Loss, we consistently outperform its original results over several different backbones. Code is available at: https://***/blisgard/BucketedRankingBasedLosses. © 2024, CC BY.

关键词： Object detection

MoCap-to-Visual Domain Adaptation for Efficient Human Mesh Estimation from 2D Keypoints

学校读者我要写书评

暂无评论

MoCap-to-Visual Domain Adaptation for Efficient Human Mesh E...

IEEE computer Society Conference on computer Vision and Pattern Recognition Workshops (CVPRW)

作者： Bedirhan Uguz Ozhan Suat Batuhan Karagoz Emre Akbas Department of Computer Engineering Middle East Technical University Ankara Turkey METU ROMER Robotics Center Middle East Technical University Ankara Turkey

ISBN: (数字)9798350365474

ISBN: (纸本)9798350365481

This paper presents Key2Mesh, a model that takes a set of 2D human pose keypoints as input and estimates the corresponding body mesh. Since this process does not involve any visual (i.e. RGB image) data, the model can be trained on large-scale motion capture (MoCap) datasets, thereby overcoming the scarcity of image datasets with 3D labels. To enable the model’s application on RGB images, we first run an off-the-shelf 2D pose estimator to obtain the 2D keypoints, and then feed these 2D keypoints to Key2Mesh. To improve the performance of our model on RGB images, we apply an adversarial domain adaptation (DA) method to bridge the gap between the MoCap and visual domains. Crucially, our DA method does not require 3D labels for visual data, which enables adaptation to target sets without the need for costly labels. We evaluate Key2Mesh for the task of estimating 3D human meshes from 2D key-points, in the absence of RGB and mesh label pairs. Our results on widely used H3.6M and 3DPW datasets show that Key2Mesh sets the new state-of-the-art by outperforming other models in PA-MPJPE for both datasets, and in MPJPE and PVE for the 3DPW dataset. Thanks to our model’s simple architecture, it operates at least 12× faster than the prior state-of-the-art model, LGD [44]. Additional qualitative samples and code are available on the project website: https://***/.

关键词： Training Adaptation models Visualization Solid modeling Three-dimensional displays Computational modeling Training data

Cascaded Model Predictive Control of Underactuated Bipedal Walking with Impact and Friction Considerations

学校读者我要写书评

暂无评论

Cascaded Model Predictive Control of Underactuated Bipedal W...

IEEE-RAS International Conference on Humanoid Robots

作者： Sait Sovukluk Christian Ott M. Mert Ankaralı Automation and Control Institute (ACIN) TU Wien Vienna Austria Department of Electrical and Electronics Engineering Middle East Technical University (METU) Ankara Turkey Institute of Robotics and Mechatronics German Aerospace Center (DLR) Wessling Germany METU Robotics and AI Technologies Application and Research Center (METU-ROMER) Ankara Turkey

This study demonstrates a cascaded model predictive control (C-MPC) method for input constrained control of underactuated planar bipedal walking with any predefined sta-bilizable trajectory. Our approach aims to increase the trajectory tracking performance of the system and produce realistic and applicable responses while respecting friction constraints and considering impact dynamics. Primarily focusing on zero dynamics with PD (ZD+PD) control, this proposed control method constitutes a second layer controller on top of the well-known trajectory tracking controllers for underactuated bipedal walking systems. We successfully implement this model-based controller and test against large modeling errors, noises, and disturbances, where the conventional ZD+PD control fails to maintain stability.

关键词：

AssembleRL: Learning to Assemble Furniture from Their Point Clouds

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Aslan, Ozgur Bolat, Burak Bal, Batuhan Tumer, Tugba Sahin, Erol Kalkan, Sinan KOVAN Research Lab METU-ROMER Center for Robotics and Artificial Intelligence Dept. of Computer Engineering Middle East Technical University Ankara Turkey

The rise of simulation environments has enabled learning-based approaches for assembly planning, which is otherwise a labor-intensive and daunting task. Assembling furniture is especially interesting since furniture are intricate and pose challenges for learning-based approaches. Surprisingly, humans can solve furniture assembly mostly given a 2D snapshot of the assembled product. Although recent years have witnessed promising learning-based approaches for furniture assembly, they assume the availability of correct connection labels for each assembly step, which are expensive to obtain in practice. In this paper, we alleviate this assumption and aim to solve furniture assembly with as little human expertise and supervision as possible. To be specific, we assume the availability of the assembled point cloud, and comparing the point cloud of the current assembly and the point cloud of the target product, obtain a novel reward signal based on two measures: Incorrectness and incompleteness. We show that our novel reward signal can train a deep network to successfully assemble different types of furniture. © 2022, CC BY-NC-SA.

关键词： Assembly

CANet: ChronoAdaptive Network for Enhanced Long-Term Time Series Forecasting under Non-Stationarity

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Sonmezer, Mert Ertekin, Seyda Department of Computer Engineering Middle East Technical University Ankara06800 Turkey METU-BILTIR CAD CAM & Robotics Research Center Middle East Technical University Ankara06800 Turkey

Long-term time series forecasting plays a pivotal role in various real-world applications. Despite recent advancements and the success of different architectures, forecasting is often challenging due to non-stationary nature of the real-world data, which frequently exhibit distribution shifts and temporal changes in statistical properties like mean and variance over time. Previous studies suggest that this inherent variability complicates forecasting, limiting the performance of many models by leading to loss of non-stationarity and resulting in over-stationarization (Liu, Wu, Wang and Long, 2022c). To address this challenge, we introduce a novel architecture, ChoronoAdaptive Network (CANet), inspired by style-transfer techniques. The core of CANet is the Non-stationary Adaptive Normalization module, seamlessly integrating the Style Blending Gate and Adaptive Instance Normalization (AdaIN) (Huang and Belongie, 2017). The Style Blending Gate preserves and reintegrates non-stationary characteristics, such as mean and standard deviation, by blending internal and external statistics, preventing over-stationarization while maintaining essential temporal dependencies. Coupled with AdaIN, which dynamically adapts the model to statistical changes, this approach enhances predictive accuracy under non-stationary conditions. CANet also employs multi-resolution patching to handle short-term fluctuations and long-term trends, along with Fourier analysis-based adaptive thresholding to reduce noise. A Stacked Kronecker Product Layer further optimizes the model’s efficiency while maintaining high performance. Extensive experiments on real-world datasets validate CANet’s superiority over state-of-the-art methods, achieving a 42% reduction in MSE and a 22% reduction in MAE. The source code is publicly available.1 © 2025, CC BY-NC-ND.

关键词： Convolutional neural networks