Search

What constitutes the "vibe" of a particular scene? What should one find in "a busy, dirty city street", "an idyllic countryside", or "a crime scene in an abandoned living room"? The translation from abstract scene descriptions to stylized scene elements cannot be done with any generality by extant systems trained on rigid and limited indoor datasets. In this paper, we propose to leverage the knowledge captured by foundation models to accomplish this translation. We present a system that can serve as a tool to generate stylized assets for 3D scenes described by a short phrase, without the need to enumerate the objects to be found within the scene or give instructions on their appearance. Additionally, it is robust to open-world concepts in a way that traditional methods trained on limited data are not, affording more creative freedom to the 3D artist. Our system demonstrates this using a foundation model "team" composed of a large language model, a vision-language model and several image diffusion models, which communicate using an interpretable and user-editable intermediate representation, thus allowing for more versatile and controllable stylized asset generation for 3D artists. We introduce novel metrics for this task, and show through human evaluations that in 91% of the cases, our system outputs are judged more faithful to the semantics of the input scene description than the baseline, thus highlighting the potential of this approach to radically accelerate the 3D content creation process for 3D artists.

BACKGROUND AND AIMS: Diabetes Mellitus (DM) is chronic conditions with devastating multi-systemic complication and may be associated with severe form of Coronavirus Disease 2019 (COVID-19). We conducted a systematic review and meta-analysis in order to investigate the association between DM and poor outcome in patients with COVID-19 pneumonia. METHODS: Systematic literature search was performed from several electronic databases on subjects that assess DM and outcome in COVID-19 pneumonia. The outcome of interest was composite poor outcome, including mortality, severe COVID-19, acute respiratory distress syndrome (ARDS), need for intensive care unit (ICU) care, and disease progression. RESULTS: There were a total of 6452 patients from 30 studies. Meta-analysis showed that DM was associated with composite poor outcome (RR 2.38 [1.88, 3.03], p

Natural language interaction is a promising direction for democratizing 3D shape design. However, existing methods for text-driven 3D shape editing face challenges in producing decoupled, local edits to 3D shapes. We address this problem by learning disentangled latent representations that ground language in 3D geometry. To this end, we propose a complementary tool set including a novel network architecture, a disentanglement loss, and a new editing procedure. Additionally, to measure edit locality, we define a new metric that we call part-wise edit precision. We show that our method outperforms existing SOTA methods by 20% in terms of edit locality, and up to 6.6% in terms of language reference resolution accuracy. Our work suggests that by solely disentangling language representations, downstream 3D shape editing can become more local to relevant parts, even if the model was never given explicit part-based supervision.

We introduce PartGlot, a neural framework and associated architectures for learning semantic part segmentation of 3D shape geometry, based solely on part referential language. We exploit the fact that linguistic descriptions of a shape can provide priors on the shape's parts -- as natural language has evolved to reflect human perception of the compositional structure of objects, essential to their recognition and use. For training, we use the paired geometry / language data collected in the ShapeGlot work for their reference game, where a speaker creates an utterance to differentiate a target shape from two distractors and the listener has to find the target based on this utterance. Our network is designed to solve this target discrimination problem, carefully incorporating a Transformer-based attention module so that the output attention can precisely highlight the semantic part or parts described in the language. Furthermore, the network operates without any direct supervision on the 3D geometry itself. Surprisingly, we further demonstrate that the learned part information is generalizable to shape classes unseen during training. Our approach opens the possibility of learning 3D shape parts from language alone, without the need for large-scale part geometry annotations, thus facilitating annotation acquisition., CVPR 2022 (Oral)

Additional file 1: Table S1. Electronic search strategy.

Supplemental material, Author_Response_1 for C-reactive protein, procalcitonin, D-dimer, and ferritin in severe coronavirus disease-2019: a meta-analysis by Ian Huang, Raymond Pranata, Michael Anthonius Lim, Amaylia Oehadian and Bachti Alisjahbana in Therapeutic Advances in Respiratory Disease

Supplemental material, Reviewer_2_v.2 for C-reactive protein, procalcitonin, D-dimer, and ferritin in severe coronavirus disease-2019: a meta-analysis by Ian Huang, Raymond Pranata, Michael Anthonius Lim, Amaylia Oehadian and Bachti Alisjahbana in Therapeutic Advances in Respiratory Disease

Supplemental material, Reviewer_2_v.1 for C-reactive protein, procalcitonin, D-dimer, and ferritin in severe coronavirus disease-2019: a meta-analysis by Ian Huang, Raymond Pranata, Michael Anthonius Lim, Amaylia Oehadian and Bachti Alisjahbana in Therapeutic Advances in Respiratory Disease

Supplemental material, Reviewer_1_v.1 for C-reactive protein, procalcitonin, D-dimer, and ferritin in severe coronavirus disease-2019: a meta-analysis by Ian Huang, Raymond Pranata, Michael Anthonius Lim, Amaylia Oehadian and Bachti Alisjahbana in Therapeutic Advances in Respiratory Disease

Supplemental material, Supplementary_Table for Multiorgan Failure With Emphasis on Acute Kidney Injury and Severity of COVID-19: Systematic Review and Meta-Analysis by Michael Anthonius Lim, Raymond Pranata, Ian Huang, Emir Yonas, Arto Yuwono Soeroto and Rudi Supriyadi in Canadian Journal of Kidney Health and Disease

Background. Metformin is widely known as an antidiabetic agent which has significant gastrointestinal side effects, but nightmares and abnormal dreams as its adverse reactions are not well reported. Case Presentation. Herein we present a case of 56-year-old male patient with no known history of recurrent nightmares and sleep disorder, experiencing nightmare and abnormal dreams directly after consumption of 750 mg extended release metformin. He reported his dream as an unpleasant experience which awakened him at night with negative feelings. The nightmare only lasted for a night, but his dreams every night thereafter seemed abnormal. The dreams were vivid and indescribable. The disappearance and occurrence of abnormal dreams ensued soon after the drug was discontinued and rechallenged. The case was assessed using Naranjo Adverse Drug Reaction (ADR) probability scale and resulted as probable causality. Conclusion. Metformin might be the underlying cause of nightmare and abnormal dreams in this patient. More studies are needed to confirm the association and causality of this findings.

Although deep learning models perform remarkably well across a range of tasks such as language translation and object recognition, it remains unclear what high-level logic, if any, they follow. Understanding this logic may lead to more transparency, better model design, and faster experimentation. Recent machine learning research has leveraged statistical methods to identify hidden units that behave (e.g., activate) similarly to human understandable logic, but those analyses require considerable manual effort. Our insight is that many of those studies follow a common analysis pattern, which we term Deep Neural Inspection. There is opportunity to provide a declarative abstraction to easily express, execute, and optimize them. This paper describes DeepBase, a system to inspect neural network behaviors through a unified interface. We model logic with user-provided hypothesis functions that annotate the data with high-level labels (e.g., part-of-speech tags, image captions). DeepBase lets users quickly identify individual or groups of units that have strong statistical dependencies with desired hypotheses. We discuss how DeepBase can express existing analyses, propose a set of simple and effective optimizations to speed up a standard Python implementation by up to 72x, and reproduce recent studies from the NLP literature.

Background. Although de Winter T-wave electrocardiography pattern is rare, it signifies proximal left anterior descending artery occlusion and is often unrecognized by physicians. The aim of this case report was to highlight the dilemma in the management of a patient with de Winter T-wave pattern in the hospital without interventional cardiology facility. Case Presentation. A 65-year-old male presented with typical chest pain since 2 hours before admission, and ECG showed sinus rhythm of 57 bpm and >1 mm upsloping ST depression with symmetric tall T in lead V2-3 characteristic of de Winter T-wave ECG pattern. He was given dual antiplatelet therapy, nitrate, statin, and anticoagulant. He refused referral to interventional cardiology available hospital. 3 hours after admission, the electrocardiography transformed into Q-waves consistent with final stages of acute STEMI and ST-segment elevation that barely meets the threshold in the guideline, and thrombolytic was administered and successful. There is a suggestion that de Winter T-wave electrocardiography should be treated as ST-segment myocardial infarction equivalent and should undergo coronary angiography; however, not every hospital has the luxury of interventional cardiology facility. The other modality for reperfusion is thrombolysis; however, without a clear guideline and scarcity of study, we prefer to resort to conservative treatment. “Fortunately,” transformation into ST-segment elevation helps us to determine the course of action which is reperfusion using thrombolytic. Conclusions. de Winter T-wave ECG pattern is not mentioned in any guidelines regarding acute coronary syndromes, and there are no clear recommendations. Physicians in rural area without interventional cardiology facility face a dilemma with the lack of evidence-based guideline. Fibrinolytic may be appropriate in those without contraindications with strong chest pain consistent with acute coronary occlusion, less than 3 hours of symptoms, and convincing de Winter T-wave ECG pattern for a rural non-PCI hospital far away from PCI capable hospital.