Zobrazeno 1 - 10
of 3 545
pro vyhledávání: '"Sanmiguel, A."'
Segmentation models are typically constrained by the categories defined during training. To address this, researchers have explored two independent approaches: adapting Vision-Language Models (VLMs) and leveraging synthetic data. However, VLMs often
Externí odkaz:
http://arxiv.org/abs/2412.09240
We consider the problem of text-to-video generation tasks with precise control for various applications such as camera movement control and video-to-video editing. Most methods tacking this problem rely on providing user-defined controls, such as bin
Externí odkaz:
http://arxiv.org/abs/2411.10501
Test-Time Adaptation for Keypoint-Based Spacecraft Pose Estimation Based on Predicted-View Synthesis
Publikováno v:
IEEE Transactions on Aerospace and Electronic Systems (2024)
Due to the difficulty of replicating the real conditions during training, supervised algorithms for spacecraft pose estimation experience a drop in performance when trained on synthetic data and applied to real operational data. To address this issue
Externí odkaz:
http://arxiv.org/abs/2410.04298
Merging parameters of multiple models has resurfaced as an effective strategy to enhance task performance and robustness, but prior work is limited by the high costs of ensemble creation and inference. In this paper, we leverage the abundance of free
Externí odkaz:
http://arxiv.org/abs/2409.15813
Autor:
Francés-Belda, Víctor, Solera-Rico, Alberto, Nieto-Centenero, Javier, Andrés, Esther, Vila, Carlos Sanmiguel, Castellanos, Rodrigo
Surrogate models that combine dimensionality reduction and regression techniques are essential to reduce the need for costly high-fidelity computational fluid dynamics data. New approaches using $\beta$-Variational Autoencoder ($\beta$-VAE) architect
Externí odkaz:
http://arxiv.org/abs/2408.04969
In unsupervised domain adaptation (UDA), where models are trained on source data (e.g., synthetic) and adapted to target data (e.g., real-world) without target annotations, addressing the challenge of significant class imbalance remains an open issue
Externí odkaz:
http://arxiv.org/abs/2407.01327
Open-Vocabulary Attention Maps with Token Optimization for Semantic Segmentation in Diffusion Models
Publikováno v:
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2024)
Diffusion models represent a new paradigm in text-to-image generation. Beyond generating high-quality images from text prompts, models such as Stable Diffusion have been successfully extended to the joint generation of semantic segmentation pseudo-ma
Externí odkaz:
http://arxiv.org/abs/2403.14291
Autor:
Yu, Xuanlong, Zuo, Yi, Wang, Zitao, Zhang, Xiaowen, Zhao, Jiaxuan, Yang, Yuting, Jiao, Licheng, Peng, Rui, Wang, Xinyi, Zhang, Junpei, Zhang, Kexin, Liu, Fang, Alcover-Couso, Roberto, SanMiguel, Juan C., Escudero-Viñolo, Marcos, Tian, Hanlin, Matsui, Kenta, Wang, Tianhao, Adan, Fahmy, Gao, Zhitong, He, Xuming, Bouniot, Quentin, Moghaddam, Hossein, Rai, Shyam Nandan, Cermelli, Fabio, Masone, Carlo, Pilzer, Andrea, Ricci, Elisa, Bursuc, Andrei, Solin, Arno, Trapp, Martin, Li, Rui, Yao, Angela, Chen, Wenlong, Simpson, Ivor, Campbell, Neill D. F., Franchi, Gianni
This paper outlines the winning solutions employed in addressing the MUAD uncertainty quantification challenge held at ICCV 2023. The challenge was centered around semantic segmentation in urban environments, with a particular focus on natural advers
Externí odkaz:
http://arxiv.org/abs/2309.15478
Autor:
Solera-Rico, Alberto, Vila, Carlos Sanmiguel, Gómez, M. A., Wang, Yuning, Almashjary, Abdulrahman, Dawson, Scott T. M., Vinuesa, Ricardo
Variational autoencoder (VAE) architectures have the potential to develop reduced-order models (ROMs) for chaotic fluid flows. We propose a method for learning compact and near-orthogonal ROMs using a combination of a $\beta$-VAE and a transformer, t
Externí odkaz:
http://arxiv.org/abs/2304.03571
In semantic segmentation, training data down-sampling is commonly performed due to limited resources, the need to adapt image size to the model input, or improve data augmentation. This down-sampling typically employs different strategies for the ima
Externí odkaz:
http://arxiv.org/abs/2302.13961