Filters








2,413 Hits in 6.5 sec

2020 Index IEEE Transactions on Cognitive and Developmental Systems Vol. 12

2020 IEEE Transactions on Cognitive and Developmental Systems  
., +, TCDS June 2020 311-322 Memory Mechanisms for Discriminative Visual Tracking Algorithms With Deep Neural Networks.  ...  ., +, TCDS Sept. 2020 439-450 Memory Mechanisms for Discriminative Visual Tracking Algorithms With Deep Neural Networks.  ...  Muscle Reducing Redundancy of Musculoskeletal Robot With Convex Hull Vertexes Selection. Zhong, S., +, TCDS Sept. 2020 601-617  ... 
doi:10.1109/tcds.2020.3044690 fatcat:yfo6c366aramfdltqegqyqphbq

Probing Physics Knowledge Using Tools from Developmental Psychology [article]

Luis Piloto, Ari Weinstein, Dhruva TB, Arun Ahuja, Mehdi Mirza, Greg Wayne, David Amos, Chia-chun Hung, Matt Botvinick
2018 arXiv   pre-print
Our results show how the VOE technique may provide a useful tool for tracking physics knowledge in future research.  ...  In order to build agents with a rich understanding of their environment, one key objective is to endow them with a grasp of intuitive physics; an ability to reason about three-dimensional objects, their  ...  This takes the form of a variational recurrent neural network (VRNN) with the Least-Recently Used (LRU) memory mechanism for memory.  ... 
arXiv:1804.01128v1 fatcat:2yxuf6ytbrhljfmjyplbcff4gy

Segmentation of Drosophila Heart in Optical Coherence Microscopy Images Using Convolutional Neural Networks [article]

Lian Duan, Xi Qin, Yuanhao He, Xialin Sang, Jinda Pan, Tao Xu, Jing Men, Rudolph E. Tanzi, Airong Li, Yutao Ma, Chao Zhou
2018 arXiv   pre-print
Convolutional neural networks are powerful tools for image segmentation and classification.  ...  With our well-trained convolutional neural network model, the heart regions through multiple heartbeat cycles can be marked with an intersection over union (IOU) of ~86%.  ...  RESULTS Prediction Results from the Neural Network The model was trained on a single NvdiaGeforce GTX 1080 GPU with 8GB memory. Each Information Video S1, S2, S3).  ... 
arXiv:1803.01947v2 fatcat:jakavlssejhuzpb5prq7fd6sge

HyperNCA: Growing Developmental Networks with Neural Cellular Automata [article]

Elias Najarro, Shyam Sudhakaran, Claire Glanois, Sebastian Risi
2022 arXiv   pre-print
In contrast to deep reinforcement learning agents, biological neural networks are grown through a self-organized developmental process.  ...  Inspired by self-organising systems and information-theoretic approaches to developmental biology, we show that our HyperNCA method can grow neural networks capable of solving common reinforcement learning  ...  Algorithm 1 : 1 HyperNCA: Growing neural networks with Neural Cellular Automata & CMAES Input: Reinforcement learning task T , NCA model, number of developmental steps δ, training hyper-parameters Ω, fitness  ... 
arXiv:2204.11674v1 fatcat:6sus3q7qyzhtdnx4eot34mqo7i

From Brain Science to Artificial Intelligence

Jingtao Fan, Lu Fang, Jiamin Wu, Yuchen Guo, Qionghai Dai
2020 Engineering  
At present, although the developmental trend in AI and its applications has surpassed expectations, an insurmountable gap remains between AI and human intelligence.  ...  toward this goal are to explore the secrets of brain science by studying new brain-imaging technology; to establish a dynamic connection diagram of the brain; and to integrate neuroscience experiments with  ...  Compliance with ethics guidelines Jingtao Fan, Lu Fang, Jiamin Wu, Yuchen Guo, and Qionghai Dai declare that they have no conflicts of interest or financial conflicts to disclose.  ... 
doi:10.1016/j.eng.2019.11.012 fatcat:qqw7gh5gpbaidhaxado4iw6ipq

Adapting the Interplay between Personalized and Generalized Affect Recognition based on an Unsupervised Neural Framework

Pablo GE Barros, Emilia Barakova, Stefan GE Wermter
2020 IEEE Transactions on Affective Computing  
In this paper, we present an unsupervised neural framework that improves emotion recognition by learning how to describe continuous affective behavior of individual persons.  ...  We evaluate our model with a series of experiments ranging from ablation studies assessing the different contributions of each neural component to an objective comparison with state-of-the-art solutions  ...  Parisi for important suggestions and support on the development of our model, and Katja Koesters for the review of this manuscript.  ... 
doi:10.1109/taffc.2020.3002657 fatcat:q7qmgkttsndhnl64ysats4xl44

EPySeg: a coding-free solution for automated segmentation of epithelia using deep learning [article]

Benoit Aigouy, Benjamin Prud'homme
2020 bioRxiv   pre-print
EPySeg, which comes with a straightforward graphical user interface, can be used as a python package on a local computer, or on the cloud via Google Colab for users not equipped with deep-learning compatible  ...  By alleviating human input in image segmentation, EPySeg accelerates and improves the characterization of epithelial tissues for all developmental biologists.  ...  This interface allows for building, training and running convolutional neural networks.  ... 
doi:10.1101/2020.06.30.179507 fatcat:xhhmptg7vrbqpbidun2yfldk6q

Continual Lifelong Learning with Neural Networks: A Review [article]

German I. Parisi, Ronald Kemker, Jose L. Part, Christopher Kanan, Stefan Wermter
2019 arXiv   pre-print
However, lifelong learning remains a long-standing challenge for machine learning and neural network models since the continual acquisition of incrementally available information from non-stationary data  ...  This limitation represents a major drawback for state-of-the-art deep neural network models that typically learn representations from stationary batches of training data, thus without accounting for situations  ...  The authors would like to thank Sascha Griffiths, Vincenzo Lomonaco, Sebastian Risi, and Jun Tani for valuable feedback and suggestions.  ... 
arXiv:1802.07569v3 fatcat:6zn2hqi2djbu3lx5mbr75nvipq

Potentials and Limitations of Deep Neural Networks for Cognitive Robots [article]

Doreen Jirak, Stefan Wermter
2018 arXiv   pre-print
Then, we identify crucial settings for cognitive robotics where deep neural networks have as yet only contributed little compared to the challenges in cognitive robotics.  ...  Although Deep Neural Networks reached remarkable performance on several benchmarks and even gained scientific publicity, they are not able to address the concept of cognition as a whole.  ...  ., CNNs is to use 3D convolution kernels for image stacks representing videos.  ... 
arXiv:1805.00777v1 fatcat:d734apeprbgaxjbom4hixrmvie

A Deep Neural Model Of Emotion Appraisal [article]

Pablo Barros, Emilia Barakova, Stefan Wermter
2018 arXiv   pre-print
In this paper, we propose a deep neural model which is designed in the light of different aspects of developmental learning of emotional concepts to provide an integrated solution for internal and external  ...  We evaluate the performance of the proposed model with different challenging corpora and compare it with state-of-the-art models for external emotion appraisal.  ...  Ekman demonstrates that facial Fig. 1 Cross-Channel Convolution Neural Network with a crossmodal architecture.  ... 
arXiv:1808.00252v1 fatcat:zp3qgbqafrdofaxbeykss6g7oy

Cellular structure image classification with small targeted training samples

Dali Wang, Zheng Lu, Yichi Xu, Zi Wang, Anthony Santella, Zhirong Bao
2019 IEEE Access  
We then transfer the structure of the GAN discriminator into a new Alex-style neural network for further learning with several dozen labeled samples.  ...  However, it is challenging to generate sufficient training samples for pattern identification through deep learning because of a limited amount of images and annotations.  ...  After initializing the neural network with parameters from the GAN discriminator, we continue to train the network using a small manually labeled dataset.  ... 
doi:10.1109/access.2019.2940161 pmid:32832309 pmcid:PMC7442139 fatcat:dmv6lsydfbbq7kmtvvxmnlyfwe

Intrinsic Motivation and Episodic Memories for Robot Exploration of High-Dimensional Sensory Spaces [article]

Guido Schillaci, Antonio Pico Villalpando, Verena Vanessa Hafner, Peter Hanappe, David Colliaux, Timothée Wintz
2020 arXiv   pre-print
A combination of deep neural networks for offline unsupervised learning of low-dimensional features from images, and of online learning of shallow neural networks representing the inverse and forward kinematics  ...  We propose the integration of an episodic memory in intrinsic motivation systems to face catastrophic forgetting issues, typically experienced when performing online updates of artificial neural networks  ...  The authors would like to thank Bruno Lara and Alejandra Ciria for the very helpful feedback on the manuscript.  ... 
arXiv:2001.01982v1 fatcat:o6hyqlqfpjatpbm7yaqbavr6ne

Understanding images in biological and computer vision

Andrew J. Schofield, Iain D. Gilchrist, Marina Bloj, Ales Leonardis, Nicola Bellotto
2018 Interface Focus  
She highlighted how such estimates might be achieved in computer vision using convolutional neural networks.  ...  Future direction: machine learning The final session of the meeting dealt with future directions for the field with an emphasis on machine learning.  ... 
doi:10.1098/rsfs.2018.0027 fatcat:xi7fvjkzozgj3pekdvh6kbcx44

If deep learning is the answer, then what is the question? [article]

Andrew Saxe, Stephanie Nelli, Christopher Summerfield
2020 arXiv   pre-print
What is the outlook for neuroscientists who seek to characterise computations or neural codes, or who wish to understand perception, attention, memory, and executive functions?  ...  Many researchers are excited by the possibility that deep neural networks may offer theories of perception, cognition and action for biological brains.  ...  Figure 3 . 3 Developmental trajectories in deep linear neural networks.  ... 
arXiv:2004.07580v2 fatcat:2ltmlfs4xbdhvhh7qcga7rcbq4

Artificial Neurogenesis: An Introduction and Selective Review [chapter]

Taras Kowaliw, Nicolas Bredeche, Sylvain Chevallier, René Doursat
2014 Studies in Computational Intelligence  
Before deep learning, most multilayered neural nets contained only one hidden layer, with the notable exception of LeCun's convolutional network [171] (see below).  ...  In convolutional networks, the weight sharing technique allows learning a specific filter for each convolution map, which drastically reduces the number of variables required, and also explains why convolutional  ... 
doi:10.1007/978-3-642-55337-0_1 fatcat:xx6nzfvbmfgzjhse6t5il3lbxe
« Previous Showing results 1 — 15 out of 2,413 results