Abstract In this article, we examine to what extent the settlement of university graduates can be derived from satellite images. We apply a convolutional neural network (CNN) to grid images of a city and predict five density classes of university graduates at a micro level (250 m × 250 m grid size). The CNN reaches an accuracy rate of 40.5% (random approach: 20%). Furthermore, the accuracy increases to 78.3% when considering a one-class deviation compared to the true class. We also examine the predictability of inhabited and uninhabited grid cells, where we achieve an accuracy of 95.3% using the same CNN. From this, we conclude that there is information that correlates with graduate density that can be derived by analysing only satellite images. The findings show the high potential of computer vision for urban and regional economics. Particularly in data-poor regions, the approach utilised facilitates comparative analytics and provides a possible solution for the modifiable aerial unit (MAU) problem. The MAU problem is a statistical bias that can influence the results of a spatial data analysis of point-estimate data that is aggregated in districts of different shapes and sizes, distorting the results.

We present a first method for the automated age estimation of buildings from unconstrained photographs. To this end, we propose a two-stage approach that firstly learns characteristic visual patterns for different building epochs at patch-level and then globally aggregates patch-level age estimates over the building. We compile evaluation datasets from different sources and perform an detailed evaluation of our approach, its sensitivity to parameters, and the capabilities of the employed deep networks to learn characteristic visual age-related patterns. Results show that our approach is able to estimate building age at a surprisingly high level that even outperforms human evaluators and thereby sets a new performance baseline. This work represents a first step towards the automated assessment of building parameters for automated price prediction.

We present an approach for the soft-output decoding in multiple-head maximum-transition-run (MTR) encoded digital magnetic recording systems. The log-likelihood ratio (LLR) algebra implementation in Boolean logic mappings has been used to propagate the soft-output information through the MTR decoder. MTR code is employed as inner code in concatenation with low-density parity-check (LDPC) in two-track two-head E2PR4 partial response (PR) magnetic recording channel. The overall decoding complexity reduction is offered by proposed simulation scheme, while nearly 1.9dB coding gain is obtained with soft-output decoding on AWGN channel.

This paper is concerned with the well-known coding technique that uses cosets of convolutional codes along with precoding to increase the free squared Euclidean distance at the output of a partial-response channel. The focus is on totally trellis-matched codes, for which the decoding trellis has the fewest number of states. Codes totally trellis-matched to the extended partial-response class 4 (EPR4) and the doubly extended partial-response class 4 (E/sup 2/PR4) channels for a variety of code rates and overall constraint lengths are sought for with respect to the squared Euclidean distance spectrum criterion. Furthermore, some previously published codes are shown to contain flawed codewords and are replaced with new ones.

The performance of maximum-likelihood decoding of trellis codes for moderate-to-high SNR is dictated by their distance spectrum. An improved version of a bidirectional algorithm is used to find cosets of convolutional codes for the EPR4 and EEPR4 channels with best (squared Euclidean) distance spectrum. Results are restricted to the class of totally trellis-matched codes. Distance spectra results for these new optimized codes as well as for some related previously published codes are tabulated. Moreover, we call attention to some codes that were erroneously tabulated in a previous paper.