Related Experiment Video
Updated: Sep 29, 2025
07:31
Defining the Role Of Language in Infants' Object Categorization with Eye-tracking Paradigms
Published on: February 8, 2019
6.7K
Learning online visual invariances for novel objects via supervised and self-supervised training.
Valerio Biscione1, Jeffrey S Bowers1
1Department of Psychology, University of Bristol, Bristol, BS8 1TL, United Kingdom.
Summary
Standard convolutional neural networks (CNNs) can achieve human-like object recognition across transformations. This study shows CNNs learn invariances with limited data and self-supervision, mimicking human visual learning.
Area of Science:
- Computer Vision
- Cognitive Science
- Artificial Intelligence
Background:
- Humans exhibit remarkable object recognition abilities despite spatial transformations like scale and viewpoint changes.
- Convolutional Neural Networks (CNNs) are models of human vision, but typically require extensive data augmentation for invariance.
- Online invariance, recognizing novel objects after single exposure, is a key human capability not fully replicated in standard CNNs.
Purpose of the Study:
- To investigate if standard CNNs can achieve human-like online invariance.
- To assess CNN performance on object recognition across various transformations (rotation, scaling, viewpoint, etc.).
- To explore self-supervised methods for acquiring invariances.
Main Methods:
- Training CNNs on synthetic 3D objects undergoing transformations.
- Analyzing internal model representations to understand invariance acquisition.
- Evaluating performance on novel classes and real-world object datasets.
- Employing a self-supervised same/different task.
Main Results:
- Standard supervised CNNs trained on transformed objects acquire strong invariances for novel classes with limited data (50 objects/10 classes).
- Invariances were transferable to a dataset of real-world object photographs.
- Self-supervised learning via a same/different task also enabled invariance acquisition.
- CNNs demonstrated robustness to rotation, scaling, translation, brightness, contrast, and viewpoint changes.
Conclusions:
- Standard CNNs can achieve human-like online invariance in object recognition.
- Limited data and self-supervision are sufficient for CNNs to learn robust invariances.
- Self-supervised approaches may mirror human visual invariance learning mechanisms.
Keywords:
Convolutional neural networksInternal representationInvariant representationOnline invarianceUnsupervised learningMore Related Videos
Related Concept Videos
01:12Observational Learning
349
Albert Bandura's observational learning, also known as imitation or modeling, occurs when a person observes and imitates another's behavior. It is a quicker process than operant conditioning. A well-known example is the Bobo doll study, where children who saw an adult acting aggressively towards the doll were more likely to act aggressively when left alone, compared to those who observed a nonaggressive adult. Many psychologists view observational learning as a form of latent learning...
349
01:18Introduction to Learning
575
Learning is the process of acquiring knowledge or skills through practice or experience, leading to long-lasting behavioral changes. This acquisition occurs through interaction with the environment and requires practice or experience. For instance, mastering a skill such as surfing requires considerable practice and experience, highlighting the essential role of repeated interactions with the environment in learning.
In contrast to learned behaviors, unlearned behaviors such as crying, sexual...
In contrast to learned behaviors, unlearned behaviors such as crying, sexual...
575
01:12Perceptual Constancy
633
Perceptual constancy is the ability to recognize that objects remain consistent and unchanged even when their appearance varies due to changes in sensory input. There are four main types of perceptual constancy: size constancy, shape constancy, color constancy, and brightness constancy.
Size constancy is the recognition that an object remains the same size, even when its image on the retina changes. For instance, a bus is perceived to be large enough to carry people, even if it looks tiny from...
Size constancy is the recognition that an object remains the same size, even when its image on the retina changes. For instance, a bus is perceived to be large enough to carry people, even if it looks tiny from...
633
01:27Associative Learning
630
Associative learning is a fundamental concept in behavioral psychology, wherein a connection is established between two stimuli or events, leading to a learned response. This process is critical in understanding how behaviors are acquired and modified. Conditioning, the mechanism through which associations are formed, can be divided into two main types: classical conditioning and operant conditioning, each elucidating different aspects of associative learning.
Classical conditioning, also known...
Classical conditioning, also known...
630
01:12Modeling and Similitude
360
Scaled modeling is a fundamental technique in engineering, enabling the study of large and complex systems by creating smaller, manageable replicas that recreate critical characteristics of the original. In hydrology and civil infrastructure, for example, scaled models of dams help analyze water flow, turbulence, and pressure. This method allows for accurate predictions of real-world behavior within a controlled environment, significantly reducing the cost and time involved in full-scale...
360