The aim of this work is learning to reshape the object in an input image to an arbitrary new shape, by just simply providing a single reference image with an object instance in the desired shape. We propose a new Generative Adversarial Network (GAN) architecture for such an object reshaping problem, named ReshapeGAN. The network can be tailored for handling all kinds of problem settings, including both within-domain (or single-dataset) reshaping and cross-domain (typically across mutiple datasets) reshaping, with paired or unpaired training data. The appearance of the input object is preserved in all cases, and thus it is still identifiable after reshaping, which has never been achieved as far as we are aware. We present the tailored models of the proposed ReshapeGAN for all the problem settings, and have them tested on 8 kinds of reshaping tasks with 13 different datasets, demonstrating the ability of ReshapeGAN on generating convincing and superior results for object reshaping. To the best of our knowledge, we are the first to be able to make one GAN framework work on all such object reshaping tasks, especially the cross-domain tasks on handling multiple diverse datasets. We present here both ablation studies on our proposed ReshapeGAN models and comparisons with the state-of-the-art models when they are made comparable, using all kinds of applicable metrics that we are aware of.

The ICDCIT 2020 proceedings are dealing with distributed computing and internet technology focusing on concurrent and distributed systems modelling and verification; cloud and grid computing; social networks, machine learning and mobile networks; data processing and blockchain technology.

Phase contrast time-lapse microscopy is a non-destructive technique that generates large volumes of image-based information to quantify the behaviour of individual cells or cell populations. To guide the development of algorithms for computer-aided cell tracking and analysis, 48 time-lapse image sequences, each spanning approximately 3.5 days, were generated with accompanying ground truths for C2C12 myoblast cells cultured under 4 different media conditions, including with fibroblast growth factor 2 (FGF2), bone morphogenetic protein 2 (BMP2), FGF2 + BMP2, and control (no growth factor). The ground truths generated contain information for tracking at least 3 parent cells and their descendants within these datasets and were validated using a two-tier system of manual curation. This comprehensive, validated dataset will be useful in advancing the development of computer-aided cell tracking algorithms and function as a benchmark, providing an invaluable opportunity to deepen our understanding of individual and population-based cell dynamics for biomedical research.

There is a natural correlation between the visual and auditive elements of a video. In this work we leverage this connection to learn general and effective models for both audio and video analysis from self-supervised temporal synchronization. We demonstrate that a calibrated curriculum learning scheme, a careful choice of negative examples, and the use of a contrastive loss are critical ingredients to obtain powerful multi-sensory representations from models optimized to discern temporal synchronization of audio-video pairs. Without further finetuning, the resulting audio features achieve performance superior or comparable to the state-of-the-art on established audio classification benchmarks (DCASE2014 and ESC-50). At the same time, our visual subnet provides a very effective initialization to improve the accuracy of video-based action recognition models: compared to learning from scratch, our self-supervised pretraining yields a remarkable gain of +19.9% in action recognition accuracy on UCF101 and a boost of +17.7% on HMDB51.

Detailed and accurate characteristics of preimplantation embryos are fundamental for a deep understanding of their development. Recent studies indicate that various geometric features of cells, such as size, shape, volume, and position play a significant role in embryo growth. However, a quantitative assessment of these characteristics first needs a segmentation of the individual cells. The manual separation and labeling of cells is extremely inefficient, and an automated approach is highly desirable. This paper presents an automatic method for early stage embryo segmentation into its constituent cells and membranes using three-dimensional (3D) data. The input data consist of two Z-stacks of fluorescence microscope images containing nuclei and membranes. The method uses a 3D level set segmentation algorithm. Its evaluation is based on a dataset composed of 20 mouse embryos, each with 4–32 blastomeres. Segmentation accuracy was evaluated by calculating F-scores with ground truth obtained by manually labeling desired regions. We also compared output of our method with the one acquired with a watershed algorithm. The proposed approach was able to achieve more than \(90\%\) accuracy for embryos with 4 and 8 cells, while for embryos with higher number of cells it was lower, reaching \(75\%\) for 32-cell embryo.

Asian rice is a staple food in Indonesia and worldwide, and its production is essential to food security. Cataloging and linking genetic variation in Asian rice to important traits, such as quality and yield, is needed in developing superior varieties of rice. We develop a bioinformatics workflow for quality control and data analysis of genetic and trait data for a diversity panel of 467 rice varieties found in Indonesia. The bioinformatics workflow operates using a back-end relational database for data storage and retrieval. Quality control and data analysis procedures are implemented and automated using the whole genome data analysis toolset, PLINK, and the [R] statistical computing language. The 467 rice varieties were genotyped using a custom array (717,312 genotypes total) and phenotyped for 12 traits in four locations in Indonesia across multiple seasons. We applied our bioinformatics workflow to these data and present prototype genome-wide association results for a continuous trait days to flowering. Two genetic variants, located on chromosome 4 and 12 of the rice genome, showed evidence for association in these data. We conclude by outlining extensions to the workflow and plans for more sophisticated statistical analyses.

Along with substantial progress made recently in designing near-optimal mechanisms for multi-item auctions, interesting structural questions have also been raised and studied. In particular, is it true that the seller can always extract more revenue from a market where the buyers value the items higher than another market? In this paper we obtain such a revenue monotonicity result in a general setting. Precisely, consider the revenue-maximizing combinatorial auction for m items and n buyers in the Bayesian setting, specified by a valuation function v and a set F of nm independent item-type distributions. Let REV (v, F) denote the maximum revenue achievable under F by any incentive compatible mechanism. Intuitively, one would expect that REV (v, G) REV (v, F) if distribution G stochastically dominates F. Surprisingly, Hart and Reny (2012) showed that this is not always true even for the simple case when v is additive. A natural question arises: Are these deviations contained within bounds? To what extent may the monotonicity intuition still be valid? We present an approximate monotonicity theorem for the class of fractionally subadditive (XOS) valuation functions v, showing that REV (v, G) cREV (v, F) if G stochastically dominates F under v where c > 0 is a universal constant. Previously, approximate monotonicity was known only for the case n = 1: Babaioff et al. (2014) for the class of additive valuations, and Rubinstein and Weinberg (2015) for all subaddtive valuation functions.

We consider scenarios where we have zero instances of real pedestrian data (e.g., a newly installed surveillance system in a novel location in which no labeled real data or unsupervised real data exists yet) and a pedestrian detector must be developed prior to any observations of pedestrians. Given a single image and auxiliary scene information in the form of camera parameters and geometric layout of the scene, our approach infers and generates a large variety of geometrically and photometrically accurate potential images of synthetic pedestrians along with purely accurate ground-truth labels through the use of computer graphics rendering engine. We first present an efficient discriminative learning method that takes these synthetic renders and generates a unique spatially-varying and geometry-preserving pedestrian appearance classifier customized for every possible location in the scene. In order to extend our approach to multi-task learning for further analysis (i.e., estimating pose and segmentation of pedestrians besides detection), we build a more generalized model employing a fully convolutional neural network architecture for multi-task learning leveraging the “free" ground-truth annotations that can be obtained from our pedestrian synthesizer. We demonstrate that when real human annotated data is scarce or non-existent, our data generation strategy can provide an excellent solution for an array of tasks for human activity analysis including detection, pose estimation and segmentation. Experimental results show that our approach (1) outperforms classical models and hybrid synthetic-real models, (2) outperforms various combinations of off-the-shelf state-of-the-art pedestrian detectors and pose estimators that are trained on real data, and (3) surprisingly, our method using purely synthetic data is able to outperform models trained on real scene-specific data when data is limited.

Recent methods for facial landmark location perform well on close-to-frontal faces but have problems in generalising to large head rotations. In order to address this issue we propose a second order linear regression method that is both compact and robust against strong rotations. We provide a closed form solution, making the method fast to train. We test the method's performance on two challenging datasets. The first has been intensely used by the community. The second has been specially generated from a well known 3D face dataset. It is considerably more challenging, including a high diversity of rotations and more samples than any other existing public dataset. The proposed method is compared against state-of-the-art approaches, including RCPR, CGPRT, LBF, CFSS, and GSDM. Results upon both datasets show that the proposed method offers state-of-the-art performance on near frontal view data, improves state-of-the-art methods on more challenging head rotation problems and keeps a compact model size.

In recent years, a rapid increase in bacterial strains resistant to modern antibiotics has been observed. This alarming rise in drug-resistant organisms has emphasized the importance of identifying new effective antimicrobial agents. Since traditional approaches for drug susceptibility testing are time-consuming and labor-intensive, more ef cient methods are urgently needed. Here, we report an automatic image analysis system for drug susceptibility testing that provides results within 3 hours using a drug susceptibility testing micro uidic (DSTM) device. The device consists of ve sets of four micro uidic channels prepared by soft lithography. The channels are in close proximity to allow simultaneous observations. The antimicrobial agent and bacterial suspension to be tested are added to the channel and incubated for 3 hours. Previously, microscopic images of the DSTM device were analyzed manually by an expert to evaluate the susceptibility of a strain. In this work, we present an automatic computer vision algorithm for processing images and performing analysis. The algorithm enhances the quality of the input image, detects cells in each channel, extracts a variety of cell-related characteristics, and estimates drug susceptibility using a pre-trained support vector machine. We address the issue of overlapping cells by incorporating a graph-based cell separation algorithm. The minimum concentration of a drug for which the proposed method predicted susceptibility represents the minimum inhibitory concentration (MIC). The novel method was implemented as a standalone application and tested on a dataset containing images of 101 clinically isolated strains of Pseudomonas aeruginosa incubated in the presence of ve different drugs. The estimated MICs correlated well with the results obtained using the conventional broth microdilution method.

We present a range of new incremental (single-pass streaming) algorithms for incremental principal components analysis (IPCA) and show that they are more effective than exiting ones. IPCA algorithms process the columns of a matrix A one at a time and attempt to build a basis for a low-dimensional subspace that spans the dominant subspace of A. We present a unified framework for IPCA algorithms, show that many existing ones are parameterizations of it, propose new sophisticated algorithms, and show that both the new algorithms and many existing ones can be implemented more efficiently than was previously known. We also show that many existing algorithms can fail even in easy cases and we show experimentally that our new algorithms outperform existing ones.

•An original “task oriented” way to categorize the state of the art of the AT works has been introduced.•The paper concentrates on a set of cross-application Computer Vision tasks that are set as the pivots to categorize AT.•The paper addresses also critical assessment of ATs considering users, medical, economical and social perspectives as well as emerging open challenges, and it tries to catch a glimpse of the possible recent advancements in Computer Vision that could allow a real improvement of the assistive outcomes.