Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Artificial Intelligence in Design

Sotra, Stefan (2023) In Diploma work IDEL01 20231
Industrial Design
Abstract
The Clip model was a major
breakthrough in the field of artificial
intelligence (AI) when it was introduced
by OpenAI in January 2021. The Clip
(Contrastive Language-Image Pre-
Training) model is a neural network
model that can understand both text and
images, enabling it to perform tasks such
as image classification and captioning, as
well as natural language processing
tasks.

The Clip model can be pre-trained on
large datasets, allowing it to learn a wide
range of patterns. This pre-training
process involves training the model on
large amounts of text and image data,
and using a contrastive learning
approach to teach it to recognize similar
patterns in both domains.

The Clip model has been used to create
large... (More)
The Clip model was a major
breakthrough in the field of artificial
intelligence (AI) when it was introduced
by OpenAI in January 2021. The Clip
(Contrastive Language-Image Pre-
Training) model is a neural network
model that can understand both text and
images, enabling it to perform tasks such
as image classification and captioning, as
well as natural language processing
tasks.

The Clip model can be pre-trained on
large datasets, allowing it to learn a wide
range of patterns. This pre-training
process involves training the model on
large amounts of text and image data,
and using a contrastive learning
approach to teach it to recognize similar
patterns in both domains.

The Clip model has been used to create
large datasets such as the Laion-5b
dataset, which contains 5.85 billion text
and image pairs. This dataset was
created by scraping the internet for
images, and then labeling them using the
Clip model. The resulting dataset is a
valuable resource for training generative
models, as it contains a large and diverse
set of text and image pairs that can be
used to train models. This in turn has
paved the way for popular image
generating AI models such as Stable Diffusion and MidJourney..

Overall, the Clip model has been a
significant breakthrough in the field of
AI, enabling the creation of large
datasets and the development of
powerful generative models. Its ability to
understand both text and images has
opened up new avenues for research and
applications. (Less)
Please use this url to cite or link to this publication:
author
Sotra, Stefan
supervisor
organization
alternative title
Roof Box Machine
course
IDEL01 20231
year
type
M2 - Bachelor Degree
subject
publication/series
Diploma work
report number
ISRN: LUT-DVIDE/EX--23/50636-SE
other publication id
ISRN
language
English
id
9130338
date added to LUP
2023-06-29 09:29:41
date last changed
2023-06-29 09:29:41
@misc{9130338,
  abstract     = {{The Clip model was a major
breakthrough in the field of artificial
intelligence (AI) when it was introduced
by OpenAI in January 2021. The Clip
(Contrastive Language-Image Pre-
Training) model is a neural network
model that can understand both text and
images, enabling it to perform tasks such
as image classification and captioning, as
well as natural language processing
tasks.

The Clip model can be pre-trained on
large datasets, allowing it to learn a wide
range of patterns. This pre-training
process involves training the model on
large amounts of text and image data,
and using a contrastive learning
approach to teach it to recognize similar
patterns in both domains.

The Clip model has been used to create
large datasets such as the Laion-5b
dataset, which contains 5.85 billion text
and image pairs. This dataset was
created by scraping the internet for
images, and then labeling them using the
Clip model. The resulting dataset is a
valuable resource for training generative
models, as it contains a large and diverse
set of text and image pairs that can be
used to train models. This in turn has
paved the way for popular image
generating AI models such as Stable Diffusion and MidJourney..

Overall, the Clip model has been a
significant breakthrough in the field of
AI, enabling the creation of large
datasets and the development of
powerful generative models. Its ability to
understand both text and images has
opened up new avenues for research and
applications.}},
  author       = {{Sotra, Stefan}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Diploma work}},
  title        = {{Artificial Intelligence in Design}},
  year         = {{2023}},
}