The issue of writer identification and writer retrieval, which is considered a challenging problem in the field of document analysis and recognition is addressed here.. A novel pipeline is proposed for the problem at hand by employing a unified neural network architecture consisting of the ResNet-20 as a feature extractor and an integrated NetV-LAD layer, inspired by the vector of locally aggregated descriptors (VLAD), in the head of the latter part. Having defined this architecture, the triplet semi-hard loss function is used to directly learn an embedding for individual input image patches. Subsequently, the generalised max-pooling technique is employed for the aggregation of embedded descriptors of each handwritten image. Also, a novel re-ranking strategy is introduced for the task of identification and retrieval based on the k-reciprocal nearest neighbours, and it is shown that the pipeline can benefit tremendously from this step. Experimental evaluation has been done on the three publicly available datasets: the ICDAR 2013, CVL, and KHATT datasets. Results indicate that while the performance was comparable to the state-of-the-art KHATT, the writer identification and writer retrieval pipeline achieve superior performance on the ICDAR 2013 and CVL datasets in terms of mAP.This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

We present an early prototype of a design system that uses Deep Learning methodology—Conditional Variational Autoencoders (CVAE)—to arrive at custom design spaces that can be interactively explored using semantic labels. Our work is closely tied to principles of parametric design. We use parametric models to create the dataset needed to train the neural network, thus tackling the problem of lacking 3D datasets needed for deep learning. We propose that the CVAE functions as a parametric tool in itself: The solution space is larger and more diverse than the combined solution spaces of all parametric models used for training. We showcase multiple methods on how this solution space can be navigated and explored, supporting explorations such as object morphing, object addition, and rudimentary 3D style transfer. As a test case, we implemented some examples of the geometric taxonomy of “Operative Design” by Di Mari and Yoo.

Building a reservoir model by integrating seismic angle stacks and well data has been challenging for oil and gas companies. Seismic data contain very detailed information of reservoir properties lateraly but lack vertical resolution while well data provide very detailed information vertically but lack horizontal information. One established method of integrating these two kinds of data is deterministic seismic inversion. Deterministic seismic inversion has proven to be a good method in delineating the reservoirs at seismic resolution but it has limitation in analyzing thin reservoirs and it only pruduced one result. Geostatistical seismic inversion method has the ability to integrate many types of data and produce multiple relaizations of the results. The realizations are at much higher details than seismic data and are able to capture thin reservoirs.An AVA geostatistical seismic inversion workflow was successfully implemented to produce highly detailed reservoir models of Abu Madi reservoir sands in Nile delta, Egypt. Abu Madi Formation is composed of lacustrine turbidite deposits in semi-isolated basin and can be subdivided into Upper and Lower reservoirs. Intraformational shale baffles occur quite commonly within Abu Madi Formation which acted as barriers of pressure depletion between the Upper and Lower reservoirs. Well data analysis has demonstrated the complex pattern of the stacked reservoir sand zones of heterogeneous reservoir parameters and pressure trends. The main objective of the study was to produce reservoir models that could be used to understand the observed pressure depletion trends within the Upper and Lower Abu Madi reservoirs which have great significance for effective field management.Seismic angle stacks and well data were integrated through AVA geostatistical inversion to produce highly detailed lithotype and elastic property results at 0.5 ms vertical sampling. Thesse models have successfully captured the shale baffles. Lithotype and elastic property realizations were used to cosimulate for reservoir (engineering) properties of Effective Porosity and Volume of Clay. The Effective Porosity and Volume of Clay realizations were then ranked to provide the P10, P50 and P90 models to be used as input for dynamic flow simulation. Five permeability rock types were derived based on extensive SCAL database and used to define the permeability and saturation models. These detailed engineering models were used for dynamic flow simulation and successfully predicted the pressure depletion trends in the Abu Madi reservoirs.

No abstract