735962 research outputs found
Sort by
Enhanced View Planning for Robotic Harvesting:Tackling Occlusions with Imitation Learning
No full textIn agricultural automation, inherent occlusion presents a major challenge for robotic harvesting. We propose a novel imitation learning-based viewpoint planning approach to actively adjust camera viewpoint and capture unobstructed images of the target crop. Traditional viewpoint planners and existing learning-based methods, depend on manually designed evaluation metrics or reward functions, often struggle to generalize to complex, unseen scenarios. Our method employs the Action Chunking with Transformer (ACT) algorithm to learn effective camera motion policies from expert demonstrations. This enables continuous six-degree-of-freedom (6-DoF) viewpoint adjustments that are smoother, more precise and reveal occluded targets. Extensive experiments in both simulated and realworld environments, featuring agricultural scenarios and a 6 DoF robot arm equipped with an RGB-D camera, demonstrate our method's superior success rate and efficiency, especially in complex occlusion conditions, as well as its ability to generalize across different crops without reprogramming. This study advances robotic harvesting by providing a practical 'learn from demonstration' (LfD) solution to occlusion challenges, ultimately enhancing autonomous harvesting performance and productivity.</p
Palate-first versus lip-first surgical repair sequence in unilateral cleft lip, alveolus, and palate:A retrospective cephalometric comparison of maxillary growth at 5-year follow-up
No full textObjective: To assess the effect of a palate-first surgical approach on maxillofacial morphology in unilateral cleft lip, alveolus, and palate (UCLP) patients compared to the conventional lip-first sequence. Subjects and Method: This retrospective cross-sectional cephalometric study included two groups of 25 non-syndromic UCLP patients. Group 1 underwent palate repair at 6–9 months, followed by lip, alveolus, and anterior palate repair with primary nasal correction after 3–6 months. Group 2 had lip, alveolus, and anterior palate repair at 3–6 months, with palate repair at 6–18 months. Lateral cephalograms taken at least five years post-second surgery were analyzed. Cephalometric variables between groups were compared using two-sample T-tests. Results: No significant differences were found in S-N-SS, PM-SS, S-PM, N-SP, S-N-Pg, NSL-ML, and ILS-NL. Only maxillary base inclination (NSL-NL) (p < 0.05) was significantly different between groups (mean difference 1.29°; p = 0.003), indicating a relatively antero-inferior maxillary tip in the palate first approach. Conclusions: In UCLP, the cleft repair protocol prioritizing palate repair at 6 months before lip repair at 9 months or later compared to conventional lip repair followed palate repair did not show significant differences in midfacial growth within the first 5 years postoperatively. However, given that pubertal growth trajectories were not captured, long-term studies are warranted to evaluate potential effects. Considering the potential to reduce treatment dropout rates in low- and middle-income countries, the palate-first approach may still be considered for broader application.</p
From Suffering to Disorders:Conceptual Analysis of Reification in Psychiatry
No full textReification means treating an abstract construct as a concrete entity. It is theorized to have negative effects in psychiatry, particularly by critiques of the biomedical model. Yet, the process of such reification receives little attention, and lack of agreement on conceptual questions prevents its quantification. The current study aims to provide a conceptual analysis of reification in psychiatry that would allow its operationalization and measurement. Problem-centered expert interviews with 10 academic and clinical experts provided insights into the nature of the reification that results in the notion of mental disorders. Thematic analysis was used to identify its core features. Reification was described as a cognitive process converting fluid human experiences into a single disorder-like entity, while suppressing their complexity and variation. The entity is then provided with its own existence independent of the subject and of the observer, with simplistic etiology and with the capacity to have causal consequences. The study identified six core features of such reification, including sub-features that may provide grounds for operationalization of the construct. The six features included structuring experiences, reducing their complexity, treating the model as reality, postulating existence of a disorder-like entity, de-subjectivization of experiences, and labeling. Reification of experiences into disorders is a series of cognitive processes affecting how individuals structure and explain mental suffering. Quantitative research is needed to further examine the effects of reification on the lived experience and prevalence of mental disorders.</p
The Uneven Impact of Big Data in Science:A Literature Review and Reflective Examination of Big Data in Data-Intensive Disciplines
No full textData practices vary widely across scientific disciplines. While Big Data has significantly transformed research activities across various domains and has been described as a revolutionary force in scientific paradigms, its application has not been uniform across all fields. This study examines Big Data research and practices in data-intensive disciplines (DIDs), identifying its distinct features and revealing the uneven adoption and impact of Big Data across scientific domains. Our findings indicate that discussions on the epistemological concepts and definitions of Big Data in DIDs are limited, with little divergence among scholars. Machine learning emerges as a central understanding and technological focus across DIDs, closely integrated with research topics and widely driving scientific advancements. Additionally, this paper highlights the instrumental role of Big Data in scientific inquiry and underscores the disparities in its impact across different disciplines. Through this review, we aim to foster a more comprehensive understanding of Big Data’s evolving role in science, emphasizing the need for continued critical reflection as its influence continues to develop
Translation and Translators:Race, Ethnicity and <i>Othello</i> in the Netherlands
No full textThis chapter looks at three translations of Othello for the Dutch stage. While the ethnic and racial backgrounds of actors playing Othello have long been an area of scholarly interest, this chapter argues that much more attention needs to be paid to the backgrounds of those who have translated the play. The author demonstrates this by situating Dutch versions of Othello by three translators (Hafid Bouazza, Jibbe Willems and Esther Duysker) in their sociopolitical contexts. The relation between translator and translation and how they mediated the minefield of identity and context was far from straightforward and any further research or debate on translation and translators in Shakespeare studies, which is still very much a tabula rasa, needs to be careful of foregoing the nuances. There are legitimate and necessary questions to be asked about the identity and experience of a translator in relation to specific Shakespeare texts, but always taking into account that reality is more complex than appearances might suggest
Structure-Preserving Approximate Balanced Reduction of Interconnected Structural-Dynamics Models
No full textThis paper considers the problem of complexity reduction of large-scale interconnected structural-dynamics models. On the one hand, traditional CMS-based reduction methods for such problems often fail to sufficiently reduce the order of these models. On the other hand, the large-scale nature of these models obstructs direct application of more effective balanced reduction methods. To address this challenge, we propose a synergetic approach that combines several structurepreserving balanced truncation methods with various efficient Gramian approximation techniques. A comparative study of the effectiveness and computational efficiency of the resulting methods is performed by using those to reduce a structuraldynamics model from the lithography industry.</p
External Test of a Deep Learning Algorithm for Pulmonary Nodule Malignancy Risk Stratification Using European Screening Data
No full textBackground: Low-dose CT screening reduces lung cancer-related deaths but has high rates of false-positive findings. A deep learning (DL) algorithm could improve nodule risk stratification but requires robust external testing.Purpose: To externally test a DL algorithm for nodule malignancy risk estimation using pooled data from three large European lung cancer screening trials.Materials and Methods: In this retrospective study, a DL algorithm trained on National Lung Screening Trial data was externally tested using baseline CT scans from the Danish Lung Cancer Screening Trial, the Multicentric Italian Lung Detection trial, and the Dutch-Belgian Lung Cancer Screening Trial. Performance was assessed across the pooled cohort and two subsets: subset A, including indeterminate nodules (5-15 mm); and subset B, including cancers size-matched to benign nodules (1:2 ratio). Performance, including the area under the receiver operating characteristic curve (AUC), was compared with the Pan-Canadian Early Detection of Lung Cancer (PanCan) model.Results: The pooled cohort included 4146 participants (median age, 58 years; 78% male participants; median smoking history, 38 pack-years) with 7614 benign and 180 malignant nodules. The DL algorithm achieved AUCs of 0.98, 0.96, and 0.94 for cancers diagnosed within 1 year, 2 years, and throughout screening, respectively, compared with 0.98, 0.94, and 0.93 ( P = .19, .02, and .46, respectively) for the PanCan model. In subset A (129 malignant and 2086 benign nodules), DL significantly outperformed PanCan across the same cancer diagnosis timeframes (respective AUCs: 0.95, 0.94, and 0.90 vs 0.91, 0.88, and 0.86; all P < .05). At 100% sensitivity for cancers diagnosed within 1 year, DL classified 68.1% of benign cases as low risk versus 47.4% for the PanCan model, a 39.4% relative reduction in false-positive findings. In subset B (180 malignant and 360 benign nodules), the AUC of the DL algorithm versus the PanCan model was 0.79 versus 0.60 ( P < .01), respectively.Conclusion: The DL algorithm outperformed the PanCan model across multiple European screening datasets, demonstrating superior malignancy prediction while substantially reducing false-positive classifications for indeterminate nodules.</p
Estimation of and y distortions in the cosmic microwave background with COBE/FIRAS data
No full textThis paper presents a novel approach to estimate the and y-distortions in the Cosmic Microwave Background (CMB) using the COBE/FIRAS data. The analysis draws from the concept of blackbody radiation inversion (BRI), a mathematical technique typically used to determine the temperature distribution from a radiated power spectrum. We study the deviations from the ideal blackbody spectrum or the spectral distortions by incorporating first a non-zero chemical potential via the Bose-Einstein distribution and then also adding the Compton parameter y while keeping the monopole temperature constant. We infer the results as probability distribution functions on these distortions. Finally, we derive and at a confidence interval. Here we show how the BRI method performs in a test-case scenario, illustrating its potential for extracting spectral distortion parameters in CMB.</p
