Controlling artificial agents from visual sensory data is an arduous task.
Recently, deep learning methods have been proposed to learn a hidden state space structure purely from data, alleviating the experimenter from this tedious design task, but resulting in an entangled, non-interpreteable state space.
When studying unconstrained behaviour and allowing mice to leave their cage to navigate a complex labyrinth, the mice exhibit foraging behaviour in the labyrinth searching for rewards, returning to their home cage now and then, e. g. to drink.
The ability of tensor networks to represent the probabilistic nature of quantum states as well as to reduce large state spaces makes tensor networks a natural candidate for active inference.
We find that our framework can generate suitable cover art for most genres, and that the visual features adapt themselves to audio feature changes.
The free energy principle, and its corollary active inference, constitute a bio-inspired theory that assumes biological agents act to remain in a restricted set of preferred states of the world, i. e., they minimize their free energy.
Finally, we also show that contrastive methods perform significantly better in the case of distractors in the environment and that our method is able to generalize goals to variations in the background.
In this paper, we propose an active inference agent that actively gathers evidence for object classifications, and can learn novel object categories over time.
Aerial navigation in GPS-denied, indoor environments, is still an open challenge.
Biologically inspired algorithms for simultaneous localization and mapping (SLAM) such as RatSLAM have been shown to yield effective and robust robot navigation in both indoor and outdoor environments.
Historically, artificial intelligence has drawn much inspiration from neuroscience to fuel advances in the field.
In case the dimensionality is not predefined, this parameter is usually determined using time- and resource-consuming cross-validation.
Active inference is a theory that underpins the way biological agent's perceive and act in the real world.
Music that is generated by recurrent neural networks often lacks a sense of direction and coherence.
Active inference is a process theory of the brain that states that all living organisms infer actions in order to minimize their (expected) free energy.
Learning to take actions based on observations is a core requirement for artificial agents to be able to be successful and robust at their task.
In this work we explore the generalization characteristics of unsupervised representation learning by leveraging disentangled VAE's to learn a useful latent space on a set of relational reasoning problems derived from Raven Progressive Matrices.
In this paper, we try to open the black box of the CNN by inspecting and visualizing the learned feature maps, in the field of dermatology.
Learning-based approaches for robotic grasping using visual sensors typically require collecting a large size dataset, either manually labeled or by many trial and errors of a robotic manipulator in the real or simulated world.
Deep neural networks require large amounts of resources which makes them hard to use on resource constrained devices such as Internet-of-things devices.
Deep residual networks (ResNets) made a recent breakthrough in deep learning.
Recurrent neural networks are nowadays successfully used in an abundance of applications, going from text, speech and image processing to recommender systems.
Binary neural networks are attractive in this case because the logical operations are very fast and efficient when implemented in hardware.
However, when learning a task using reinforcement learning, the agent cannot distinguish the characteristics of the environment from those of the task.
Deep reinforcement learning is becoming increasingly popular for robot control algorithms, with the aim for a robot to self-learn useful feature representations from unstructured sensory input leading to the optimal actuation policy.
Traditional textual representations, such as tf-idf, have difficulty grasping the semantic meaning of such texts, which is important in applications such as event detection, opinion mining, news recommendation, etc.
In this paper we propose a technique which avoids the evaluation of certain convolutional filters in a deep neural network.
We present four training and prediction schedules from the same character-level recurrent neural network.
We therefore investigated several text representations as a combination of word embeddings in the context of semantic pair matching.