Embodied neuromorphic intelligence
The design of robots that interact autonomously with the environment and exhibit complex
behaviours is an open challenge that can benefit from understanding what makes living …
behaviours is an open challenge that can benefit from understanding what makes living …
Neuromorphic computing hardware and neural architectures for robotics
Neuromorphic hardware enables fast and power-efficient neural network–based artificial
intelligence that is well suited to solving robotic tasks. Neuromorphic algorithms can be …
intelligence that is well suited to solving robotic tasks. Neuromorphic algorithms can be …
Advancing neuromorphic computing with loihi: A survey of results and outlook
Deep artificial neural networks apply principles of the brain's information processing that led
to breakthroughs in machine learning spanning many problem domains. Neuromorphic …
to breakthroughs in machine learning spanning many problem domains. Neuromorphic …
Latent replay for real-time continual learning
Training deep neural networks at the edge on light computational devices, embedded
systems and robotic platforms is nowadays very challenging. Continual learning techniques …
systems and robotic platforms is nowadays very challenging. Continual learning techniques …
[PDF][PDF] Rehearsal-Free Continual Learning over Small Non-IID Batches.
Robotic vision is a field where continual learning can play a significant role. An embodied
agent operating in a complex environment subject to frequent and unpredictable changes is …
agent operating in a complex environment subject to frequent and unpredictable changes is …
CVPR 2020 continual learning in computer vision competition: Approaches, results, current challenges and future directions
In the last few years, we have witnessed a renewed and fast-growing interest in continual
learning with deep neural networks with the shared objective of making current AI systems …
learning with deep neural networks with the shared objective of making current AI systems …
Uncertainty for identifying open-set errors in visual object detection
Deployed into an open world, object detectors are prone to open-set errors, false positive
detections of object classes not present in the training dataset. We propose GMM-Det, a real …
detections of object classes not present in the training dataset. We propose GMM-Det, a real …
What's in the black box? the false negative mechanisms inside object detectors
In object detection, false negatives arise when a detector fails to detect a target object. To
understand why object detectors produce false negatives, we identify five 'false negative …
understand why object detectors produce false negatives, we identify five 'false negative …
RGB-D-based object recognition using multimodal convolutional neural networks: a survey
Object recognition in real-world environments is one of the fundamental and key tasks in
computer vision and robotics communities. With the advanced sensing technologies and low …
computer vision and robotics communities. With the advanced sensing technologies and low …
[HTML][HTML] Top-tuning: A study on transfer learning for an efficient alternative to fine tuning for image classification with fast kernel methods
The impressive performance of deep learning architectures is associated with a massive
increase in model complexity. Millions of parameters need to be tuned, with training and …
increase in model complexity. Millions of parameters need to be tuned, with training and …