Slap swarm algorithm with memory mechanism and boundary collision processing

Functional organic field-effect transistors (OFETs) have developed rapidly, especially OFETs with memory function. We make a comprehensive summary of the background, memory mechanism, structure construction and memory applications based on OFETs.

Download Full-text

A combination use of bagging and random subspace with memory mechanism for dynamic financial distress prediction

2016 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM) ◽

10.1109/ieem.2016.7798055 ◽

2016 ◽

Author(s):

Chong Wu ◽

Jiaming Liu

Keyword(s):

Financial Distress ◽

Random Subspace ◽

Memory Mechanism ◽

Financial Distress Prediction ◽

Distress Prediction ◽

With Memory

Download Full-text

MemCap: Memorizing Style Knowledge for Image Captioning

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v34i07.6998 ◽

2020 ◽

Vol 34 (07) ◽

pp. 12984-12992 ◽

Cited By ~ 2

Author(s):

Wentian Zhao ◽

Xinxiao Wu ◽

Xiaoxun Zhang

Keyword(s):

Language Model ◽

Attention Mechanism ◽

Visual Content ◽

Image Captioning ◽

Memory Module ◽

Training Corpus ◽

Memory Mechanism ◽

Linguistic Style ◽

Related Part ◽

With Memory

Generating stylized captions for images is a challenging task since it requires not only describing the content of the image accurately but also expressing the desired linguistic style appropriately. In this paper, we propose MemCap, a novel stylized image captioning method that explicitly encodes the knowledge about linguistic styles with memory mechanism. Rather than relying heavily on a language model to capture style factors in existing methods, our method resorts to memorizing stylized elements learned from training corpus. Particularly, we design a memory module that comprises a set of embedding vectors for encoding style-related phrases in training corpus. To acquire the style-related phrases, we develop a sentence decomposing algorithm that splits a stylized sentence into a style-related part that reflects the linguistic style and a content-related part that contains the visual content. When generating captions, our MemCap first extracts content-relevant style knowledge from the memory module via an attention mechanism and then incorporates the extracted knowledge into a language model. Extensive experiments on two stylized image captioning datasets (SentiCap and FlickrStyle10K) demonstrate the effectiveness of our method.

Download Full-text

An Approach to Scene Recognition Based on the Local Interaction of a Group of Robots

Mekhatronika Avtomatizatsiya Upravlenie ◽

10.17587/mau.22.94-103 ◽

2021 ◽

Vol 22 (2) ◽

pp. 94-103

Author(s):

A. D. Moskovsky ◽

E. V. Burgov ◽

E. E. Ovsyannikova

Keyword(s):

Visual Memory ◽

Recognition Algorithm ◽

Scene Recognition ◽

Local Interaction ◽

Maximum Efficiency ◽

Visual Analyzer ◽

Memory Mechanism ◽

A Value ◽

Time Period ◽

With Memory

The paper considers a group of animates (bio-inspired abstract models or technical devices) that use the visual analyzer described by the authors earlier. The visual analyzer recognizes the scene observed by the animat, determines the class of the situation and selects the corresponding behavior. The analyzer is based on the" language of poses" , borrowed from some species of ants, which allows them to visually inform other individuals of their condition. The first proposed algorithm describes a visual memory mechanism that allows an animat to remember seen objects for some time and select its behavior in a more stable way compared to a basic visual analyzer. The second algorithm describes a group recognition approach. Robots are able to exchange information about the scene they are observing presented in the form of a graph using local communication. The robot, receiving such information from its neighbors, is able to expand its picture of the world and more correctly choose its behavior based on the recognized situation than if it only had its own data. A demonstration task was set to test the efficiency, approximately simulating a colony of ants that gather food in a certain area. Colony, nest, food, and enemy individuals are modeled in the developed simulation environment. Group efficiency was defined as the amount of food collected over the allotted time period. Animates using the visual memory algorithm showed an increase in efficiency for all the studied memory parameters. Moreover, the dependence is not monotonic, and there is a certain value at which maximum efficiency is achieved. The use of the group recognition algorithm also showed an increase in efficiency compared to the basic visual analyzer. The experiments were carried out for different distances of local interaction. As with memory, there is a value of interaction distance at which the maximum efficiency is achieved. The combination of both algorithms also has interesting results. The experiments showed that the smaller the memory of the robot, the greater the increase in efficiency from using the collective recognition algorithm.

Download Full-text