Magnus Johnsson

Magnus Johnsson AI Research AB

I propose a number of principles that I believe are substantial for various faculties of the mammalian brain, such as perception, expectations, imagery, and memory. The same principles are also of interest when designing an artificial but biologically inspired cognitive architecture. More-over, I discuss how the same principles may lie behind the a...

Human recognition of the actions of other humans is very efficient and is based on patterns of movements. Our theoretical starting point is that the dynamics of the joint movements is important to action categorization. On the basis of this theory, we present a novel action recognition system that employs a hierarchy of Self-Organizing Maps togethe...

We present an online system for real time recognition of actions involving objects working in online mode. The system merges two streams of information processing running in parallel. One is carried out by a hierarchical self-organizing map (SOM) system that recognizes the performed actions by analysing the spatial trajectories of the agent's movem...

There is a growing awareness that the complexity of managing Big Data is one of the main challenges in the developing field of the Internet of Things (IoT). Complexity arises from several aspects of the Big Data life cycle, such as gathering data, storing them onto cloud servers, cleaning and integrating the data, a process involving the last advan...

We present an architecture able to recognise pitches and to internally simulate likely continuations of partially heard melodies. Our architecture consists of a novel version of the Associative Self-Organizing Map (A-SOM) with generalized ancillary connections. We tested the performance of our architecture with melodies from a publicly available da...

The new sensing applications need enhanced computing capabilities to handle the requirements of complex and huge data processing. The Internet of Things (IoT) concept brings processing and communication features to devices. In addition, the Cloud Computing paradigm provides resources and infrastructures for performing the computations and outsourci...

Human recognition of the actions of other humans is very efficient and is based on patterns of movements. Our theoretical starting point is that the dynamics of the joint movements is important to action categorization. On the basis of this theory, we present a novel action recognition system that employs a hierarchy of Self-Organizing Maps togethe...

We present a novel action recognition system that is able to learn how to recognize and classify actions. Our system employs a three-layered neural network hierarchy consisting of two self-organizing maps together with a supervised neural network for labeling the actions. The system is equipped with a module that pre- processes the 3D input data be...

We present a hierarchical self-organizing map based system for online recognition of human actions. We have made a first evaluation of our system by training it on two different sets of recorded human actions, one set containing manner actions and one set containing result actions, and then tested it by letting a human performer carry out the actio...

We propose a hierarchical neural architecture able to recognise observed human actions. Each layer in the architecture represents increasingly complex human activity features. The first layer consists of a SOM which performs dimensionality reduction and clustering of the feature space. It represents the dynamics of the stream of posture frames in a...

We present and evaluate a novel supervised recurrent neural network architecture, the SARASOM, based on the associative self-organizing map. The performance of the SARASOM is evaluated and compared with the Elman network as well as with a hidden Markov model (HMM) in a number of prediction tasks using sequences of letters, including some experiment...

We propose a system able to represent others' actions as well as to internally simulate their likely continuation from a partial observation. The approach presented here is the first step towards a more ambitious goal of endowing an artificial agent with the ability to recognise and predict others' intentions. Our approach is based on the associati...

We propose a hierarchical neural architecture able to recognise observed human actions. Each layer in the architecture represents increasingly complex human activity features. The first layer consists of a SOM which performs dimensionality reduction and clustering of the feature space. It represents the dynamics of the stream of posture frames in a...

When artificial agents interact and cooperate with other agents, either human or artificial, they need to recognize others' actions and infer their hidden intentions from the sole observation of their surface level movements. Indeed, action and intention understanding in humans is believed to facilitate a number of social interactions and is suppor...

Fertility rates have dramatically decreased in the last two decades, especially in men. It has been described that environmental factors as well as life habits may affect semen quality. In this paper we use artificial intelligence techniques in order to predict semen characteristics from environmental factors, life habits and health status, as a po...

We present a system that can learn to represent actions as well as to internally simulate the likely continuation of their initial parts. The method we propose is based on the Associative Self Organizing Map (A-SOM), a variant of the Self Organizing Map. By emulating the way the human brain is thought to perform pattern recognition tasks, the A- SO...

We present the Associative Self-Organizing Map (A-SOM) and propose that it could be used to predict an agent's intentions by internally simulating the behaviour likely to follow initial movements. The A-SOM is a neural network that develops a representation of its input space without supervision, while simultaneously learning to associate its activ...

Several recent works deal with 3D data in mobile robotic problems, e.g. mapping or egomotion. Data comes from any kind of sensor such as stereo vision systems, time of flight cameras or 3D lasers, providing a huge amount of unorganized 3D data. In this paper, we describe an efficient method to build complete 3D models from a Growing Neural Gas (GNG...

Urinary incontinence is a considerable problem which is clearly reflected in the number of patients affected by it. Moreover, it is extremely difficult to obtain an accurate diagnosis as the urinary incontinence very often is related to the neurological system. In this article a model with capabilities for urological diagnosing is proposed. This mo...

Several recent works deal with 3D data in mobile robotic problems, e.g. mapping. Data come from any kind of sensor (time of flight cameras and 3D lasers) providing a huge amount of unorganized 3D data. In this paper we detail an efficient method to build complete 3D models from a Growing Neural Gas (GNG). We show that the use of GNG provides better...

The classical connectionist models are not well suited to working with data varying over time. According to this, temporal connectionist models have emerged and constitute a continuously growing research field. In this paper we present a novel supervised recurrent neural network architecture (SARASOM) based on the Associative Self-Organizing Map (A...

We present experiments with a multimodal system based on a novel variant of the Self-Organizing Map (SOM) called the Associative Self-Organizing Map (A-SOM). The A-SOM is similar to the SOM and develops a representation of its input space, but also learns to associate its activity with additional inputs, e.g. the activities of one or several extern...

Simulation hypothesis, Cognitive modelling. Abstract: We present a study of a novel variant of the Self-Organizing Map (SOM) called the Associative Self- Organizing Map (A-SOM). The A-SOM is similar to the SOM and thus develops a representation of its input space, but in addition it also learns to associate its activi ty with the activity of one or...

We present a bimodal model able to internally simulate expected sequences of perceptions within a modality likely to follow the last sensory experience. Simultaneously reasonable perceptual expectations are elicited in the other modality. Our model is based on the novel associative self-organizing map (A-SOM), which develops a representation of the...

We present a study of neural network architectures able to internally simulate perceptions and actions. All these architectures employ the novel Associative Self-Organizing Map (A-SOM) as a perceptual neural network. The A-SOM develops a representation of its input space, but in addition also learns to associate its activity with an arbitrary numbe...

Urinary incontinence is one of the largest diseases affecting between 10% and 30% of the adult population and an increase is expected in the next decade with rising treatment costs as a consequence. There are many types of urological dysfunctions causing urinary incontinence, which makes cheap and accurate diagnosing an important issue. This paper...

The Ikaros project started in 2001 with the aim of developing an open infrastructure for system-level brain modeling. The system has developed into a general tool for cognitive modeling as well as robot control. Here we describe the main parts of the Ikaros system and how it has been used to implement various cognitive systems and to control a numb...

In this article a fuzzy system with capabilities for urological diagnosing is proposed. This system is specialized towards the diagnosis of urological dysfunctions with neurological etiology. For this reason the system specifies all the neural centres involved in both the urological phases, voiding and micturition. The fuzzy system allows to classi...

We have implemented and compared four biologically motivated self-organizing haptic systems based on proprioception. All systems employ a 12-d.o.f. anthropomorphic robot hand, the LUCS Haptic Hand 3. The four systems differ in the kind of self-organizing neural network used for clustering. For the mapping of the explored objects, one system uses a...

In this article, we evaluate the work out of some artificial neural network models as tools for support in the medical diagnosis of urological dysfunctions. We develop two types of unsupervised and one supervised neural network. This scheme is meant to help the urologists in obtaining a diagnosis for complex multi-variable diseases and to reduce pa...

We present a study of a novel variant of the Self-Organizing Map (SOM) called the Associative Self-Organizing Map (A-SOM). The A-SOM is similar to the SOM and thus develops a representation of its input space, but in addition it also learns to associate its activity with the activity of one or several external SOMs. The A-SOM has relevance in e.g....

A self-organizing neural network is described that can associate between different modalities and also has the ability to learn perceptual sequences. This architecture is a step towards the development of a complete agent containing simplified versions of all major neural subsystems in a mammal. It aims at exploring as well as takes inspiration fro...

We have experimented with different neural network based architectures for bio-inspired self-organizing texture and hardness perception systems. To this end we have developed a microphone based texture sensor and a hardness sensor that measures the compression of the material at a constant pressure.We have implemented and successfully tested both m...

We have implemented four bio-inspired self-organizing haptic systems based on proprioception on a 12 d.o.f. anthropomorphic robot hand. The four systems differ in the kind of self-organizing neural network used for clustering. For the mapping of the explored objects, one system uses a Self-Organizing Map (SOM), one uses a Growing Cell Structure (GC...

In this article we evaluate the work out of artificial neural networks as tools for helping and support in the medical diagnosis. In particular we compare the usability of one supervised and two unsupervised neural network architectures for medical diagnoses of lower urinary tract dysfunctions. The purpose is to develop a system that aid urologists...

Three different models of tactile shape perception inspired by the human haptic system were tested using an 8 d.o.f. robot hand with 45 tactile sensors. One model is based on the tensor product of different proprioceptive and tactile signals and a self-organizing map (SOM). The two other models replace the tensor product operation with a novel self...

The Lucs Haptic Hand III has been built as a step in a project at LUCS aiming at studying haptic perception. In this project, several robot hands together with cog- nitive computational models of the corresponding hu- man neurophysiologic systems will be built. Grasping tests with the LUCS Haptic Hand III were done with six different objects in ord...

We have experimented with proprioception in a bio-inspired self-organizing haptic system. To this end a 12 d.o.f. anthropomorphic robot hand with proprioceptive sensors was developed. The system uses a self-organizing map for the mapping of the explored objects. In our experiments the system was trained and tested with 10 different objects of diffe...

We have developed an 8 d.o.f. robot hand which has been tested with three computational models of haptic perception. Two of the models are based on the tensor product of different proprioceptive and sensory signals and a self-organizing map (SOM), and one uses a novel self-organizing neural network, the T-MPSOM. The computational models have been t...

We have developed an 8 d.o.f. robot hand which has been tested with three computational models of haptic perception. Two of the models are based on the tensor product of different proprioceptive and sensory signals and a self-organizing map (SOM), and one uses a novel self-organizing neural network, the T-MPSOM. The computational models have been t...

This paper describes a system for haptic object categorization. It consists of a robotic hand, the LUCS Haptic Hand I, together with software modules that to some extent simulate the functioning of the primary and the secondary somatosensory cortices. The haptic system is the first one in a project at LUCS aiming at studying haptic perception. In t...

This paper describes a robotic hand, LUCS Haptic Hand I, that has been built as a first step in a project aiming at the study haptic perception. Grasping tests with the hand were done with different objects, and the signal patterns from the sensors were studied and an- alyzed. The results suggest that LUCS Haptic Hand I provides signal patterns tha...

We have experimented with a bio-inspired self-organizing texture and hardness perception system which automatically learns to associate the representations of the two submodalities with each other. To this end we have developed a microphone based texture sensor and a hardness sensor that measures the compression of the material at a constant pressu...

We have developed an 8 d.o.f. robot hand, which has been tested with two computational models of haptic perception. Each model uses a variant of the novel self-organizing neural net- work, the Tensor-Multiple Peak Self-Organizing Map (T-MPSOM). One of the models uses a vari- ant of the T-MPSOM that multiplies the activ- ity corresponding to each of...

We describe a number of research projects at Lund Univer- sity Cognitive Science. The first project focuses on visual attention. The second area is haptic perceptions by robots. The third area is anticipation in groups of robots. The final research area concerns building computa- tional tools that can be used to design large-scale cognitive models...

We present a study of supervised neural network architectures capable of internal simulation of perceptions and actions. These architectures employ the novel Associative Self-Organizing Map (A-SOM) as a hidden layer (for the representation of perceptions), and a neural network adapted by the delta rule as an output layer (for the representation of...

After nerve injuries, e.g. following a reimplantation of an accidentally cut hand, nerve fibres grows together again but will be misconnected. This implies an incorrectly mapping of the hand in somatosensory cortex. However, the brain is plastic and thus a natural question is if there is a way to stimulate the hand of such a patient so that the rec...