MuTr: Multi-Stage Transformer for Hand Pose Estimation from Full-Scene Depth Image.

This work presents a novel transformer-based method for hand pose estimation-DePOTR. We test the DePOTR method on four benchmark datasets, where DePOTR outperforms other transformer-based methods while achieving results on par with other state-of-the-art methods. To further demonstrate the strength of DePOTR, we propose a novel multi-stage approach from full-scene depth image-MuTr. MuTr removes the necessity of having two different models in the hand pose estimation pipeline-one for hand localization and one for pose estimation-while maintaining promising results. To the best of our knowledge, this is the first successful attempt to use the same model architecture in standard and simultaneously in full-scene image setup while achieving competitive results in both of them. On the NYU dataset, DePOTR and MuTr reach precision equal to 7.85 mm and 8.71 mm, respectively.

One Model is Not Enough: Ensembles for Isolated Sign Language Recognition.

In this paper, we dive into sign language recognition, focusing on the recognition of isolated signs. The task is defined as a classification problem, where a sequence of frames (i.e., images) is recognized as one of the given sign language glosses. We analyze two appearance-based approaches, I3D and TimeSformer, and one pose-based approach, SPOTER. The appearance-based approaches are trained on a few different data modalities, whereas the performance of SPOTER is evaluated on different types of preprocessing. All the methods are tested on two publicly available datasets: AUTSL and WLASL300. We experiment with ensemble techniques to achieve new state-of-the-art results of 73.84% accuracy on the WLASL300 dataset by using the CMA-ES optimization method to find the best ensemble weight parameters. Furthermore, we present an ensembling technique based on the Transformer model, which we call Neural Ensembler.

Why Use Position Features in Liver Segmentation Performed by Convolutional Neural Network.

Liver volumetry is an important tool in clinical practice. The calculation of liver volume is primarily based on Computed Tomography. Unfortunately, automatic segmentation algorithms based on handcrafted features tend to leak segmented objects into surrounding tissues like the heart or the spleen. Currently, convolutional neural networks are widely used in various applications of computer vision including image segmentation, while providing very promising results. In our work, we utilize robustly segmentable structures like the spine, body surface, and sagittal plane. They are used as key points for position estimation inside the body. The signed distance fields derived from these structures are calculated and used as an additional channel on the input of our convolutional neural network, to be more specific U-Net, which is widely used in medical image segmentation tasks. Our work shows that this additional position information improves the results of the segmentation. We test our approach in two experiments on two public datasets of Computed Tomography images. To evaluate the results, we use the Accuracy, the Hausdorff distance, and the Dice coefficient. Code is publicly available at: https://gitlab.com/hachaf/liver-segmentation.git.

Semantic Segmentation of Intralobular and Extralobular Tissue from Liver Scaffold H&E Images.

Decellularized tissue is an important source for biological tissue engineering. Evaluation of the quality of decellularized tissue is performed using scanned images of hematoxylin-eosin stained (H&E) tissue sections and is usually dependent on the observer. The first step in creating a tool for the assessment of the quality of the liver scaffold without observer bias is the automatic segmentation of the whole slide image into three classes: the background, intralobular area, and extralobular area. Such segmentation enables to perform the texture analysis in the intralobular area of the liver scaffold, which is crucial part in the recellularization procedure. Existing semi-automatic methods for general segmentation (i.e., thresholding, watershed, etc.) do not meet the quality requirements. Moreover, there are no methods available to solve this task automatically. Given the low amount of training data, we proposed a two-stage method. The first stage is based on classification of simple hand-crafted descriptors of the pixels and their neighborhoods. This method is trained on partially annotated data. Its outputs are used for training of the second-stage approach, which is based on a convolutional neural network (CNN). Our architecture inspired by U-Net reaches very promising results, despite a very low amount of the training data. We provide qualitative and quantitative data for both stages. With the best training setup, we reach 90.70% recognition accuracy.