Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 47 for “"DETR"”.

  1. ROMAn-DETR: We came, We sawed, We built a Robust Orientation-aware Model with Anisotropic attention

    … To solve this we introduce our model ROMAn-DETR, an instance segmentation model made to handle anisotropic objects while also improving robustness for outof- distribution data. By combining geometric priors for these anisotropic objects, together with a new auxiliary orientation module …

    chalmers Repository record for ROMAn-DETR: We came, We sawed, We built a Robust Orientation-aware Model with Anisotropic attention (opens in a new tab)

  2. Self-Supervised Fixed-Scene Adaptation for Object-Detection in Real-Time Surveillance: A Comparative Study of YOLO11 and RF-DETR

    … architectures: the Transformer-dominant RF-DETR and the convolutional neural network (CNN)- dominant YOLO11. The thesis employs the 100Scenes dataset to represent a broad range of surveillance environments. Experimental results demonstrate that RF-DETR consistently achieves higher accuracy, …

    chalmers Repository record for Self-Supervised Fixed-Scene Adaptation for Object-Detection in Real-Time Surveillance: A Comparative Study of YOLO11 and RF-DETR (opens in a new tab)

  3. Longitudinal Resistance Training in the Elderly: Effects of3 Years of De-training on the Retention of Strength

    … (TR), ceased to weight train after 2 years (DETR), or had acted as controls throughout (CON). The TR and DETR trained hard (progressing up to 3 sets at up to 80% of I RM) for 2 years; the TR continued training for an additional 3 years at a maintenance level (2-3 sets at 60-70% 1RM), whereas …

    mcmaster Repository record for Longitudinal Resistance Training in the Elderly: Effects of3 Years of De-training on the Retention of Strength (opens in a new tab)

  4. Tag-based image retrieval using pre-trained object detection models

    … multiple state-of-the-art models including YOLO, DETR, Vision Transformer (ViT), DETR Panoptic, and Grounding DINO, the system extracts visually grounded tags that enable efficient and accurate image retrieval. Unlike embedding-based similarity or caption-based approaches that may suffer from …

    texas Repository record for Tag-based image retrieval using pre-trained object detection models (opens in a new tab)

  5. Modèle basé sur le Transformer pour une détection robuste du visage des nourrissons et des enfants hospitalisés en utilisant des images RVB et thermiques

    … sur les Transformers, DEtection TRansformer (Detr), préentraîné sur l’ensemble des donnees WiderFace, pour détecter les visages en USIP. Nous avons également utilisé les images thermiques pour améliorer la précision de la détection faciale. Nos résultats montrent que le modèle Detr se …

    ets-quebec Download PDF for Modèle basé sur le Transformer pour une détection robuste du visage des nourrissons et des enfants hospitalisés en utilisant des images RVB et thermiques (opens in a new tab)

  6. Automated Distress Detection for Asphalt and Continuously Reinforced Concrete Pavements in Inland and Coastal Regions

    … object detection architectures, YOLOv9-C and RT-DETR-L, were trained and evaluated using fused two-dimensional (2D) intensity and three-dimensional (3D) range pavement images captured by the Highway Monitoring and Inspection Tool (HyMIT) HY3D4K system. The primary objective was to determine how …

    texas-state Repository record for Automated Distress Detection for Asphalt and Continuously Reinforced Concrete Pavements in Inland and Coastal Regions (opens in a new tab)

  7. ROBUST MULTI-OBJECT DETECTION UNDER LOW-LIGHT CONDITIONS: A COMPARATIVE STUDY OF TRANSFORMER-BASED DETECTORS WITH LOW-LIGHT AUGMENTED FINE-TUNING

    … object detection architectures — DETR (DEtection TRansformer), YOLOS (You Only Look at One Sequence), and RT-DETR (Real-Time DEtection TRansformer) — evaluated under both normal and degraded lighting conditions. We propose a three-phase experimental methodology: (1) pretrained …

    csusb Repository record for ROBUST MULTI-OBJECT DETECTION UNDER LOW-LIGHT CONDITIONS: A COMPARATIVE STUDY OF TRANSFORMER-BASED DETECTORS WITH LOW-LIGHT AUGMENTED FINE-TUNING (opens in a new tab)

  8. Optimalizace neuronové sítě pro detekci objektů s využitím NPU

    … neuronových sítí, se zaměřením na sítě YOLOv8 a DETR. Tyto sítě bylo možné v lepším případě kvantizovat se ztrátou přesnosti 0,4, 0,3 a 0,6 mAP v případě verzí n, s a m modelu YOLOv8 a 3,5 mAP v případě modelu DETR-ResNet-50. Použitým datasetem byl COCO val2017. Práce zahrnuje také …

    brno-tech Repository record for Optimalizace neuronové sítě pro detekci objektů s využitím NPU (opens in a new tab)

  9. Advancing utility pole and sign detection through deep learning

    … (Faster R-CNN), and Detection Transformer (DETR). The DETR-based approach achieved mean averageprecision (mAP) values of 90.43% for pole detection and 88.26% for warning-signdetection. Third, it extends the DETR-based framework with segmentation capability, enabling pole masks to be used for …

    strathclyde Repository record for Advancing utility pole and sign detection through deep learning (opens in a new tab)

  10. Towards Precise Localization in Object Detection: Stable Matching Strategies and Advanced Bounding Box Regression Loss Functions in Deep Neural Networks

    … matching in detection transformers, such as DETR, which streamline object detection by eliminating anchor generation and non-maximum suppression (NMS). Unlike NMS's one-to-many matching, DETR employs the Hungarian algorithm for unique assignments between predictions and ground truth, …

    unsw Repository record for Towards Precise Localization in Object Detection: Stable Matching Strategies and Advanced Bounding Box Regression Loss Functions in Deep Neural Networks (opens in a new tab)

  11. Tramba: A Hybrid Architecture for Table Understanding

    … Mamba encoder with the Detection Transformer (DETR) framework to address the tasks of table detection and structure recognition. Leveraging Mamba’s state space modeling, which reduces computational complexity from O(N^2) to O(N), the proposed architecture retains competitive representational …

    concordia Download PDF for Tramba: A Hybrid Architecture for Table Understanding (opens in a new tab)

  12. UAV-Based Deep Learning for Multi-Crop Plant Counting, Geolocation, and Field-Scale Decision Support

    … evaluation procedures to compare YOLO-style and DETR-style object detectors under varying crop morphology, plant density, background texture, and illumination conditions. The study benchmarks YOLOv12, YOLO26, RT-DETR, and RF-DETR using detection-level metrics and count-level agreement metrics. …

    vt Repository record for UAV-Based Deep Learning for Multi-Crop Plant Counting, Geolocation, and Field-Scale Decision Support (opens in a new tab)

  13. Artificial Intelligence Techniques for Target Detection and Tracking From Multi Imagery Sensors

    … detection approach that com-bines YOLOv7 and DETR. This integration effectively balances local object detection precision with global context awareness, resulting in significant improvements in object localization accu-racy. An automatic selection module, based on mean average precision (mAP) …

    calgary Repository record for Artificial Intelligence Techniques for Target Detection and Tracking From Multi Imagery Sensors (opens in a new tab)

  14. Development and analysis of a small-scale controlled dataset with various weather conditions, lighting, and route types for autonomous driving

    … to generate ground truth data via the Co-DETR 2D prediction algorithm. We validated the accuracy of the Co-DETR algorithm through partial manual annotation. However, it is undeniable that in some extreme conditions, the algorithm-generated ground truth can lead to results deviating from …

    vt Repository record for Development and analysis of a small-scale controlled dataset with various weather conditions, lighting, and route types for autonomous driving (opens in a new tab)

  15. RRPN Data-Centric: Potencializando a fusão radar-câmera com supervisão 2D de alta fidelidade

    … geradas automaticamente pelo detector RF-DETR. Além disso, a arquitetura original da RRPN foi modernizada por meio da migração de frameworks deprecados para realizar a integração de backbones convolucionais modernos, como ResNeXt, WideResNet e EfficientNet. Os experimentos realizados no …

    brazil-ufpe Repository record for RRPN Data-Centric: Potencializando a fusão radar-câmera com supervisão 2D de alta fidelidade (opens in a new tab)

  16. Detecção e segmentação automática de defeitos em pás de aerogeradores utilizando modelos da família YOLO

    … F1-score. As a complementary reference, the RF-DETR-seg-m model, which is based on a different architectural paradigm, is also evaluated. Among the YOLO models, YOLOv8s achieves the best overall performance for segmentation masks, with an mAP@0.5 of 0.653, an mAP@0.5:0.95 of 0.319, a recall of …

    brazil-ufrn Repository record for Detecção e segmentação automática de defeitos em pás de aerogeradores utilizando modelos da família YOLO (opens in a new tab)

  17. Understanding the relationship between community involvement and health : Is social capital important?

    … to involve communities within decision making (DETR, 1998; 2000a). Community involvement, and the subsequent creation of networks, has been presented within national policy documentation as a means to tackle social exclusion and associated multiple deprivation (DETR, 2000a; SEU, 2001). …

    sheffield-hallam Download PDF for Understanding the relationship between community involvement and health : Is social capital important? (opens in a new tab)

  18. Robustness Investigation, Detection, and Defense of Deep learning Models against False Data Injection

    … evaluation of Detection Transformer (DETR) robustness against adversarial attacks. We extend three classic adversarial attacks FGSM, PGD, and C\&W to the object detection domain and conduct extensive experiments on MS COCO and KITTI datasets. Our findings reveal that DETR models …

    clemson Repository record for Robustness Investigation, Detection, and Defense of Deep learning Models against False Data Injection (opens in a new tab)

Page 1 of 3