Search

The use of deep learning methods for precision farming is gaining increasing interest. However, collecting training data in this application field is particularly challenging and costly due to the need of acquiring information during the different growing stages of the cultivation of interest. In this paper, we present a method for data augmentation that uses two GANs to create artificial images to augment the training data. To obtain a higher image quality, instead of re-creating the entire scene, we take original images and replace only the patches containing objects of interest with artificial ones containing new objects with different shapes and styles. In doing this, we take into account both the foreground (i.e., crop samples) and the background (i.e., the soil) of the patches. Quantitative experiments, conducted on publicly available datasets, demonstrate the effectiveness of the proposed approach. The source code and data discussed in this work are available as open source. • GAN-based augmentation enhances crop detection. • Synthetic data boosts segmentation performance. • DCGAN & cGAN improve realism in synthetic data. • We use shape & style augmentation. • Our method supports multispectral data. [ABSTRACT FROM AUTHOR]

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.robot.2006.06.007 Byline: Giorgio Grisetti (a)(b), Gian Diego Tipaldi (b), Cyrill Stachniss (c)(a), Wolfram Burgard (a), Daniele Nardi (b) Abstract: Rao-Blackwellized particle filters have become a popular tool to solve the simultaneous localization and mapping problem. This technique applies a particle filter in which each particle carries an individual map of the environment. Accordingly, a key issue is to reduce the number of particles and/or to make use of compact map representations. This paper presents an approximative but highly efficient approach to mapping with Rao-Blackwellized particle filters. Moreover, it provides a compact map model. A key advantage is that the individual particles can share large parts of the model of the environment. Furthermore, they are able to reuse an already computed proposal distribution. Both techniques substantially speed up the overall filtering process and reduce the memory requirements. Experimental results obtained with mobile robots in large-scale indoor environments and based on published standard datasets illustrate the advantages of our methods over previous mapping approaches using Rao-Blackwellized particle filters. Author Affiliation: (a) University of Freiburg, Department of Computer Science, D-79110 Freiburg, Germany (b) Dipartimento Informatica e Sistemistica, Universita 'La Sapienza', I-00198 Rome, Italy (c) Eidgenossische Technische Hochschule Zurich (ETH), IRIS, 8092 Zurich, Switzerland Article History: Received 1 October 2005; Revised 1 April 2006; Accepted 1 June 2006

A study focusing on various aspects of coordination in the framework of the RoboCup competitions, particularly addressing both multi-agent systems that are developed within the simulation league and multi-robot systems, which are realized in the middle-size league, is presented. The conclusions show that in both simulated and robotics scenarios, a major issue in coordination is to find a suitable balance between the robot individual capabilities and the form of cooperation realized.

Abstract: Using the same procedures as Nardi and Bingman (2009) , bilateral hippocampal lesions were found to have no detectable effect on the capacity of homing pigeons to use the slope of an inclined surface to encode a goal location. Hippocampal lesioned pigeons, like controls, also preferentially relied on slope over geometry when the two sources of information were set in conflict. As such, slope resembles visual features as a source of goal recognition information that is hippocampal independent. [Copyright &y& Elsevier]

Abstract: Control, right and left HF lesioned homing pigeons (Columba livia) were trained to locate a goal in one corner of a rectangular enclosure with a distinctive feature cue. Probe tests revealed that all groups were able to encode in parallel geometric (enclosure shape) and feature information, and in the absence of one of them, they could us the other to locate the goal. However, left HF lesioned pigeons learned the task at a faster rate, and when the geometric and feature information were set in conflict, they relied more on the feature cue compared to control and right HF lesioned pigeons. It was also found that pigeons, independent of group, trained to a goal adjacent to the feature cue learned the task in fewer sessions and relied more on feature information compared to pigeons trained to a goal opposite the feature cue. The latter group relied more on geometric information. The results support the hypothesis that the left HF plays a more important role in the representation of a goal location with respect to environmental shape/geometry. We further propose that the observed functional asymmetry can be explained by the lateralized properties of the pigeon tectofugal visual system. [Copyright &y& Elsevier]

The relationship between hippocampal aging and spatial-cognitive decline in birds has recently been investigated. However, like its mammalian counterpart, the avian hippocampus does not work in isolation and its relationship to the septum is of particular interest. The current study aimed to investigate the effects of age on septum (medial and lateral) and associated nucleus of the diagonal band (NDB) neuronal activation (as indicated by c-Fos expression) during learning of a spatial, delayed non-match-to-sample task conducted in a modified radial arm maze. The results indicated significantly reduced septum, but not NDB, activation during spatial learning in older pigeons. We also preliminarily investigated the effect of age on the number of cholinergic septum and NDB neurons (as indicated by expression of choline acetyltransferase; ChAT). Although underpowered to reveal a statistical effect, the data suggest that older pigeons have substantially fewer ChAT-expressing cells in the septum compared to younger pigeons. The data support the hypothesis that reduced activation of the septum contributes to the age-related, spatial cognitive impairment in pigeons. [ABSTRACT FROM AUTHOR]

Developing automatic diagnostic tools for the early detection of skin cancer lesions in dermoscopic images can help to reduce melanoma-induced mortality. Image segmentation is a key step in the automated skin lesion diagnosis pipeline. In this paper, a fast and fully-automatic algorithm for skin lesion segmentation in dermoscopic images is presented. Delaunay Triangulation is used to extract a binary mask of the lesion region, without the need of any training stage. A quantitative experimental evaluation has been conducted on a publicly available database, by taking into account six well-known state-of-the-art segmentation methods for comparison. The results of the experimental analysis demonstrate that the proposed approach is highly accurate when dealing with benign lesions, while the segmentation accuracy significantly decreases when melanoma images are processed. This behavior led us to consider geometrical and color features extracted from the binary masks generated by our algorithm for classification, achieving promising results for melanoma detection. [ABSTRACT FROM AUTHOR]

Abstract: We present a methodology for Data Warehouse design and its application within the Telecom Italia information system. The methodology is based on a conceptual representation of the Enterprise, which is exploited both in the integration phase of the Warehouse information sources and during the knowledge discovery activity on the information stored in the Warehouse. The application of the methodology in the Telecom Italia framework has been supported by prototype software tools both for conceptual modeling and for data integration and reconciliation. [Copyright &y& Elsevier]

Efficient robotic behaviors require robustness and adaptation to dynamic changes of the environment, whose characteristics rapidly vary during robot operation. To generate effective robot action policies, planning and learning techniques have shown the most promising results. However, if considered individually, they present different limitations. Planning techniques lack generalization among similar states and require experts to define behavioral routines at different levels of abstraction. Conversely, learning methods usually require a considerable number of training samples and iterations of the algorithm. To overcome these issues, and to efficiently generate robot behaviors, we introduce LoOP , an iterative learning algorithm for optimistic planning that combines state-of-the-art planning and learning techniques to generate action policies. The main contribution of LoOP is the combination of Monte-Carlo Search Planning and Q-learning, which enables focused exploration during policy refinement in different robotic applications. We demonstrate the robustness and flexibility of LoOP in various domains and multiple robotic platforms, by validating the proposed approach with an extensive experimental evaluation. • Novel approach to take advantage of both Robot learning and planning techniques. • Improve sample efficiency for tackling robotic tasks with large state-spaces. • Action policy generalization via Monte Carlo tree search and function approximation. • Extensive experimental evaluation of the proposed method in robotic scenarios. [ABSTRACT FROM AUTHOR]

Multidimensional times series prediction is a challenging task. Only recently the increased data availability has made it possible to tackle with such problems. In this work we devised a novel method to exploit the multiple correlated features in the time series. The recurrent neural networks and the wavelet transform have been important innovations in the fields of signal processing and time series prediction. This paper proposes a Wavelet Recurrent Network for multi-steps ahead prediction of multidimensional time series. The proposed model combines these two elements into a neural network that predicts multiple samples in the future that are multiple time steps ahead with respect to the input samples. This Wavelet Recurrent Network carries out a multiresolution decomposition of the input signal through the wavelet transform, predicts the future wavelet coefficients with the recurrent neural network and transforms the output back in the time domain. The proposed model is applied to the prediction of satellite telemetry data, that is composed of readings from multiple sensors which are highly correlated. The prediction of such telemetries can help the engineers to detect anomalies in the system, that, in the context of space missions, are particularly dangerous since they can compromise the entire mission if not handled properly. The results show that the proposed model outperforms the recurrent network without wavelet transform both in terms of accuracy and in the width of the forecast horizon. • The model combines RNN and the Wavelet Transform into a unique neural network. • Exploits the multiple correlated features of satellite telemetries for prediction. • Overcomes traditional analytical and numerical models, that are more expensive. • Yields multi-steps ahead medium-term forecasts with an high accuracy. • Outperforms the simple RNN in terms of accuracy and width of the forecast horizon. [ABSTRACT FROM AUTHOR]

An effective perception system is a fundamental component for farming robots, as it enables them to properly perceive the surrounding environment and to carry out targeted operations. The most recent methods make use of state-of-the-art machine learning techniques to learn a valid model for the target task. However, those techniques need a large amount of labeled data for training. A recent approach to deal with this issue is data augmentation through Generative Adversarial Networks (GANs), where entire synthetic scenes are added to the training data, thus enlarging and diversifying their informative content. In this work, we propose an alternative solution with respect to the common data augmentation methods, applying it to the fundamental problem of crop/weed segmentation in precision farming. Starting from real images, we create semi-artificial samples by replacing the most relevant object classes (i.e., crop and weeds) with their synthesized counterparts. To do that, we employ a conditional GAN (cGAN), where the generative model is trained by conditioning the shape of the generated object. Moreover, in addition to RGB data, we take into account also near-infrared (NIR) information, generating four channel multi-spectral synthetic images. Quantitative experiments, carried out on three publicly available datasets, show that (i) our model is capable of generating realistic multi-spectral images of plants and (ii) the usage of such synthetic images in the training process improves the segmentation performance of state-of-the-art semantic segmentation convolutional networks. • Novel and effective data augmentation technique, applied to the crop/weed segmentation problem in agricultural robotics. • Starting from real images, we create semi-artificial samples by replacing the most relevant object classes with synthesized counterparts. • We employ a conditional GAN, where the generative model is trained by conditioning the shape of the generated objects. • In addition to RGB data, we take into account also near-infrared information, generating 4-channels multi-spectral synthetic images. • New pixel-wise labeled dataset made publicly available with this paper. [ABSTRACT FROM AUTHOR]