Papers with Code · CC BY-SA 4.0

데이터셋 탐색

데이터셋의 설명·홈페이지를 찾고, 관련 태스크의 평가 기록을 살펴보세요.

관련 논문 9편

CHAOS - Combined (CT-MR) Healthy Abdominal Organ Segmentation

CHAOS
CHAOS 설명 보기

CHAOS challenge aims the segmentation of abdominal organs (liver, kidneys and spleen) from CT and MRI data. ONsite section of the CHAOS was held in The IEEE International Symposium on Biomedical Imaging (ISBI) on April 11, 2019, Venice, ITALY. Online submissions are still welcome! \textbf{Challenge Description} Understanding prerequisites of complicated medical procedures plays an important role in the success of the operations. To enrich the level of understanding, physicians use advanced tools such as three-dimensional visualization and printing, which require extraction of the object(s) of interest from DICOM images. Accordingly, the precise segmentation of abdominal organs (i.e. liver, kidney(s) and spleen) has critical importance for several clinical procedures including but not limited to pre-evaluation of liver for living donor-based transplantation surgery or detailed analysis of abdominal organs to determine the vessels arising from and entering them for correct positioning of a graft prior to abdominal aortic surgery. This motivates ongoing research to achieve better segmentation results and overcoming countless challenges originating from both highly flexible anatomical properties of abdomen and limitations of modalities reflected to image characteristics. In this context, the proposed challenge has two separate but related aims: 1) Segmentation of liver from computed tomography (CT) data sets, which are acquired at portal phase after contrast agent injection for pre-evaluation of living donated liver transplantation donors. 2) Segmentation of four abdominal organs (i.e. liver, spleen, right and left kidneys) from magnetic resonance imaging (MRI) data sets acquired with two different sequences (T1-DUAL and T2-SPIR). CHAOS tasks contain combination of these organs' segmentation. \textbf{Tasks} There are five competition categories in which the participating teams can take place and submit their result(s): 1) Liver Segmentation (CT & MRI): This is also called "cross-modality" [1] and it is simply based on using a single system, which can segment liver from both CT and MRI. For instance, the training and test sets of a machine learning approach would have images from both modalities without explicitly feeding the model with corresponding information. A unique study about this is a reference below and this task is one of the most interesting tasks of the challenge. Keep in mind that the fusion of individual systems for different modalities (i.e. two models, one working on CT and the other on MRI ) would not be valid for this category. They can be evaluated as individual systems at Tasks 2 and 3. On the other hand, in this task, fusion of individual systems between MR sequences (i.e. two models, one working on T1-DUAL and the other on T2-SPIR ) is allowed. 2) Liver Segmentation (CT only): This is mostly a regular task of liver segmentation from CT, (such as SLIVER07). This task is easier than SLIVER07 as it only contains healthy livers aligned in the same direction and patient position. However, the challenging part is the enhanced vascular structures (portal phase) due to the contrast injection. 3) Liver Segmentation (MRI only): Similar to "Task 2", this is also a regular task of liver segmentation from MRI. It includes two different pulse sequences: T1-DUAL and T2-SPIR. Moreover, T1-DUAL has two forms (in and out phase). The developed system should work on both sequences. In this task, the fusion of individual systems between MR sequences (i.e. two models, one working on T1-DUAL and the other on T2-SPIR ) are allowed. 4) Segmentation of abdominal organs (CT & MRI): This task is extension of Task 1 to kidneys and spleen in MRI data. In this task, the interesting part is that CT datasets have only liver, but the MRI datasets have four annotated abdominal organs (liver, kidneys, spleen). Keep in mind that fusion of individual systems for different modalities (i.e. two models, one working on CT and the other on MRI ) would not be valid for this category. On the other hand, in this task, fusion of individual systems between MR sequences (i.e. two models, one working on T1-DUAL and the other on T2-SPIR ) are allowed. 5) Segmentation of abdominal organs (MRI only): The same task given in "Task 3" but extended to four abdominal organs; liver, kidneys, spleen. In this task, ensemble or fusion of individual systems between MR sequences (i.e. two models, one working on T1-DUAL and the other on T2-SPIR ) are allowed. [1] Valindria, V. et al. (2018, March). Multi-modal learning from unpaired images: Application to multi-organ segmentation in CT and MRI. In 2018 IEEE Winter Conference on Applications of Computer Vision (WACV) (pp. 547-556). IEEE. https://doi.ieeecomputersociety.org/10.1109/WACV.2018.00066

관련 논문 27편

ChaosNLI

ChaosNLI 설명 보기

Chaos NLI is a Natural Language Inference (NLI) dataset with 100 annotations per example (for a total of 464,500 annotations) for some existing data points in the development sets of SNLI, MNLI, and Abductive NLI. The dataset provides additional labels for NLI annotations that reflect the distribution of human annotators, instead of picking the majority label as the gold standard label. Source: ChaosNLI Github Repository

관련 논문 1편

CausalChaos!QA

CausalChaos!
CausalChaos! 설명 보기

CausalChaos! is a dataset for causal video question answering. It is based on Tom and Jerry cartoons. It features longer causal chains embedded in dynamic visual scenes. It also features challenging incorrect options, especially, Causal Confusion set which contains causally confounding incorrect options. All these factors prove to be challenging for current VLMs and other traditional Video Question Answering models.

관련 논문 1편

ChaosBench

ChaosBench 설명 보기

We propose ChaosBench, a large-scale, multi-channel, physics-based benchmark for subseasonal-to-seasonal (S2S) climate prediction. It is framed as a high-dimensional video regression task that consists of 45-year, 60-channel observations for validating physics-based and data-driven models, and training the latter. Physics-based forecasts are generated from 4 national weather agencies with 44-day lead-time and serve as baselines to data-driven forecasts. Our benchmark is one of the first to incorporate physics-based metrics to ensure physically-consistent and explainable models. We establish two tasks: full and sparse dynamics prediction. 🔗: https://leap-stc.github.io/ChaosBench/ 📚: https://arxiv.org/abs/2402.00712 ## Getting Started Step 1: Clone the ChaosBench Github repository Step 2: Install package dependencies cd ChaosBench pip install -r requirements.txt Step 3: Initialize the data space by running cd data/ wget https://huggingface.co/datasets/LEAP/ChaosBench/resolve/main/process.sh chmod +x process.sh Step 5: Download the data # NOTE: you can also run each line one at a time to retrieve individual dataset ./process.sh era5 # Required: For input ERA5 data ./process.sh climatology # Required: For climatology ./process.sh ukmo # Optional: For simulation from UKMO ./process.sh ncep # Optional: For simulation from NCEP ./process.sh cma # Optional: For simulation from CMA ./process.sh ecmwf # Optional: For simulation from ECMWF