Deep learning has been successfully showing promising results in plant disease detection, fruit counting, yield estimation, and gaining an increasing interest in agriculture. Deep learning models are generally based on several millions of parameters that generate exceptionally large weight matrices. The latter requires large memory and computational power for training, testing, and deploying. Unfortunately, these requirements make it difficult to deploy on low-cost devices with limited resources that are present at the fieldwork. In addition, the lack or the bad quality of connectivity in farms does not allow remote computation. An approach that has been used to save memory and speed up the processing is to compress the models. In this work, we tackle the challenges related to the resource limitation by compressing some state-of-the-art models very often used in image classification. For this we apply model pruning and quantization to LeNet5, VGG16, and AlexNet. Original and compressed models were applied to the benchmark of plant seedling classification (V2 Plant Seedlings Dataset) and Flavia database. Results reveal that it is possible to compress the size of these models by a factor of 38 and to reduce the FLOPs of VGG16 by a factor of 99 without considerable loss of accuracy.

Wikipedia is one of the main sources of information on the Web. But the access to this content may be difficult especially when using a basic telephone without browsing capability and only a GSM network. The only means of text-based communication remains through SMS. Due to the limitation of the number of characters, a Wikipedia page cannot always be sent through SMS. This work raises the issue of text summarization with character limitation. To solve this issue, two extractive approaches have been combined: LSA and TextRank algorithms. Generated summaries have been evaluated using ROUGE metrics. Since ROUGE metrics do not consider character limitation, a new threshold named Threshold of Acceptability for Character-Oriented Summaries (TACOS) has been proposed to appreciate ROUGE metrics. The evaluation showed the relevance of the approach for pages of at most 2000 characters. The system has been tested using the SMS simulator of RapidSMS without a GSM gateway to simulate the deployment in a real environment. To the best of our knowledge, this is the first work tackling text summarization issue with character limitation.

Wireless mesh networks appear as an appealing solution to reduce the digital divide between rural and urban regions. However the placement of router nodes is still a critical issue when planning this type of network, especially in rural regions where we usually observe low density and sparse population. In this paper, we firstly provide a network model tied to rural regions by considering the area to cover as decomposed into a set of elementary areas which can be required or optional in terms of coverage and where a node can be placed or not. Afterwards, we try to determine an optimal number and positions of mesh router nodes while maximizing the coverage of areas of interest, minimizing the coverage of optional areas and ensuring connectivity of all mesh router nodes. For that we propose a particularized algorithm based on Metropolis approach to ensure an optimal coverage and connectivity with an optimal number of routers. The proposed algorithm is evaluated on different region instances. We obtained a required coverage between 94% and 97% and a coverage percentage of optional areas less than 16% with an optimal number of routers nr max-2 =1.3*nr min , (nr min being the minimum number of router which is the ratio between the total area requiring coverage and the area which can be covered by a router).professor and the chair of the Mathematics and Sciences Department, Texas A&M -Central Texas (Killeen,TX). His Current research interests include algorithm design and optimization with applications to communications systems and epidemiology.

This paper provides a deep evaluation of the energy consumption of routing protocols. The evaluation is done along with other metrics such as throughput and packet delivery ratio (PDR). We introduce two more metrics to capture the efficiency of the energy consumption: e-throughput and e-PDR. Both are ratios in relation to the energy. We consider the three low layers of the stack. Three types of routing protocols are used: proactive, reactive, and hybrid. At the MAC and PHY layer, three radio types are considered: 802.11a/b/g. Finally, the number of nodes is varying in random topologies, with nodes being static or mobile. Simulations are conducted using NS3. The parameters of a real network interface card are used. From the results in mobile position scenarios, no protocol is outperforming the others; even if OLSR has the lowest energy consumption, most of the time. However, in constant position scenarios, AODV consumed a lower energy, apart from the scenarios using the 802.11a standard where HWMP energy consumption is the lowest. Regarding the energy efficiency, AODV protocols provided the best e-throughput and OLSR the best e-PDR in overall configurations. A framework for selecting energy-efficient routing protocol depending on network characteristics is proposed at the end.

No abstract