79 research outputs found
Revisiting Fully Convolutional Geometric Features for Object 6D Pose Estimation
No full textRecent works on 6D object pose estimation focus on learning keypoint correspondences between images and object models, and then determine the object pose through RANSAC-based algorithms or by directly regressing the pose with end-to-end optimisations. We argue that learning point-level discriminative features is overlooked in the literature. To this end, we revisit Fully Convolutional Geometric Features (FCGF) and tailor it for object 6D pose estimation to achieve state-of-the-art performance. FCGF employs sparse convolutions and learns point-level features using a fully-convolutional network by optimising a hardest contrastive loss. We can outperform recent competitors on popular benchmarks by adopting key modifications to the loss and to the input data representations, by carefully tuning the training strategies, and by employing data augmentations suitable for the underlying problem. We carry out a thorough ablation to study the contribution of each modification
Open-vocabulary object 6D pose estimation
No full textWe introduce the new setting of open-vocabulary object 6D pose estimation, in which a textual prompt is used to specify the object of interest. In contrast to existing approaches, in our setting (i) the object of interest is specified solely through the textual prompt, (ii) no object model (e.g., CAD or video sequence) is required at inference, and (iii) the object is imaged from two RGBD viewpoints of different scenes. To operate in this setting, we introduce a novel approach that leverages a Vision-Language Model to segment the object of interest from the scenes and to estimate its relative 6D pose. The key of our approach is a carefully devised strategy to fuse object-level information provided by the prompt with local image features, resulting in a feature space that can generalize to novel concepts. We validate our approach on a new benchmark based on two popular datasets, REAL275 and Toyota-Light, which collectively encompass 34 object instances appearing in four thousand image pairs. The results demonstrate that our approach outperforms both a well-established handcrafted method and a recent deep learning-based baseline in estimating the relative 6D pose of objects in different scenes. Code and dataset are available at https://jcorsetti.github.io/oryon.LIDIA
Functionality understanding and segmentation in 3D scenes
No full textUnderstanding functionalities in 3D scenes involves interpreting natural language descriptions to locate functional interactive objects, such as handles and buttons, in a 3D environment. Functionality understanding is highly challenging, as it requires both world knowledge to interpret language and spatial perception to identify fine-grained objects. For example, given a task like 'turn on the ceiling light', an embodied AI agent must infer that it needs to locate the light switch, even though the switch is not explicitly mentioned in the task description. To date, no dedicated methods have been developed for this problem. In this paper, we introduce Fun3DU, the first approach designed for functionality understanding in 3D scenes. Fun3DU uses a language model to parse the task description through Chain-of-Thought reasoning in order to identify the object of interest. The identified object is segmented across multiple views of the captured scene by using a vision and language model. The segmentation results from each view are lifted in 3D and aggregated into the point cloud using geometric information. Fun3DU is training-free, relying entirely on pre-trained models. We evaluate Fun3DU on SceneFun3D, the most recent and only dataset to benchmark this task, which comprises over 3000 task descriptions on 230 scenes. Our method significantly outperforms state-of-the-art open-vocabulary 3D segmentation approaches
High-Resolution Open-Vocabulary Object 6D Pose Estimation
No full text: The generalisation to unseen objects in the 6D pose estimation task is very challenging. While Vision-Language Models (VLMs) enable using natural language descriptions to support 6D pose estimation of unseen objects, these solutions underperform compared to model-based methods. In this work we present Horyon, an open-vocabulary VLM-based architecture that addresses relative pose estimation between two scenes of an unseen object, described by a textual prompt only. We use the textual prompt to identify the unseen object in the scenes and then obtain high-resolution multi-scale features. These features are used to extract cross-scene matches for registration. We evaluate our model on a benchmark with a large variety of unseen objects across four datasets, namely REAL275, Toyota-Light, Linemod, and YCB-Video. Our method achieves state-of-the-art performance on all datasets, outperforming by 12.6 in Average Recall the previous best-performing approach
Data in support of a central role of plasminogen activator inhibitor-2 polymorphism in recurrent cardiovascular disease risk in the setting of high HDL cholesterol and C-reactive protein using Bayesian network modeling
No full textAbstractData is presented that was utilized as the basis for Bayesian network modeling of influence pathways focusing on the central role of a polymorphism of plasminogen activator inhibitor-2 (PAI-2) on recurrent cardiovascular disease risk in patients with high levels of HDL cholesterol and C-reactive protein (CRP) as a marker of inflammation, “Influences on Plasminogen Activator Inhibitor-2 Polymorphism-Associated Recurrent Cardiovascular Disease Risk in Patients with High HDL Cholesterol and Inflammation” (Corsetti et al., 2016; [1]). The data consist of occurrence of recurrent coronary events in 166 post myocardial infarction patients along with 1. clinical data on gender, race, age, and body mass index; 2. blood level data on 17 biomarkers; and 3. genotype data on 53 presumptive CVD-related single nucleotide polymorphisms. Additionally, a flow diagram of the Bayesian modeling procedure is presented along with Bayesian network subgraphs (root nodes to outcome events) utilized as the data from which PAI-2 associated influence pathways were derived (Corsetti et al., 2016; [1])
História geral do Rio Grande do Sul
No full textResenha dos livros:BOEIRA, Nelson; GOLIN, Tau (Coordenação Geral). História Geral do Rio Grande do Sul. 5 volumes. Passo Fundo: Méritos, 2006/2009. Vol. 1. Colônia. “O ensino nas crônicas do Prof. Coruja” [309- 322]. Ana Inez Klein. Vol. 2. Império. “A Instrução” [449-489]. Jaime Giolo. Vol. 3. República Velha. “A Educação: construindo o cidadão” [287-311]. Berenice Corsetti. Vol. 4. República (1930-1985). “A Educação” [315-333]. Elomar Tambara; Claudemir de Quadros; Maria Helena Camara Bastos. “Educação Superior (1930-1985)” [335-354]. Clarissa Eckert Baeta Neves.</p
Recent advances in the characterization of gaseous and liquid fuels by vibrational spectroscopy
No full textDate of Acceptance: 20/04/2015 Acknowledgments The author would like to thank Thomas Seeger, Alfred Leipertz, Florian Zehentbauer, Stella Corsetti, David McGloin, and Kristina Noack for fruitful discussions over the past decade. Special thanks to Lynda Cromwell and Andrew Williamson for proofreading the manuscriptPeer reviewe
Revisiting Fully Convolutional Geometric Features for Object 6D Pose Estimation
No full textRecent works on 6D object pose estimation focus on learning keypoint
correspondences between images and object models, and then determine the object
pose through RANSAC-based algorithms or by directly regressing the pose with
end-to-end optimisations. We argue that learning point-level discriminative
features is overlooked in the literature. To this end, we revisit Fully
Convolutional Geometric Features (FCGF) and tailor it for object 6D pose
estimation to achieve state-of-the-art performance. FCGF employs sparse
convolutions and learns point-level features using a fully-convolutional
network by optimising a hardest contrastive loss. We can outperform recent
competitors on popular benchmarks by adopting key modifications to the loss and
to the input data representations, by carefully tuning the training strategies,
and by employing data augmentations suitable for the underlying problem. We
carry out a thorough ablation to study the contribution of each modification.
The code is available at https://github.com/jcorsetti/FCGF6D.Comment: Camera ready version, 18 pages and 13 figures. Published at the 8th
International Workshop on Recovering 6D Object Pos
Functional abdominal cramping pain: Expert practical guidance
No full textFunctional abdominal cramping pain (FACP) is a common complaint, which may present either on its own or in association with a functional gastrointestinal disorder. It is likely caused by a variety of, probably partly unknown, etiologies. Effective management of FACP can be challenging owing to the lack of usable diagnostic tools and the availability of a diverse range of treatment approaches. Practical guidance for their selection and use is limited. The objective of this article is to present a working definition of FACP based on expert consensus, and to propose practical strategies for the diagnosis and management of this condition for physicians, pharmacists, and patients. A panel of experts on functional gastrointestinal disorders was convened to participate in workshop activities aimed at defining FACP and agreeing upon a recommended sequence of diagnostic criteria and management recommendations. The key principles forming the foundation of the definition of FACP and suggested management algorithms include the primacy of cramping pain as the distinguishing symptom; the importance of recognizing and acting upon alarm signals of potential structural disease; the recognition of known causes that might be addressed through lifestyle adjustment; and the central role of antispasmodics in the treatment of FACP. The proposed algorithm is intended to assist physicians in reaching a meaningful diagnostic endpoint based on patient-reported symptoms of FACP. We also discuss how this algorithm may be adapted for use by pharmacists and patients. Copyright © 2022 The Author(s). Published by Wolters Kluwer Health, Inc.https://doi.org/10.1097/mcg.000000000000176
- …
