60 lines
1.4 KiB
Markdown
60 lines
1.4 KiB
Markdown
# CSE559A Lecture 1
|
|
|
|
## Introducing the syllabus
|
|
|
|
See the syllabus on Canvas.
|
|
|
|
## Motivational introduction for computer vision
|
|
|
|
Computer vision is the study of manipulating images.
|
|
|
|
Automatic understanding of images and videos
|
|
|
|
1. vision for measurement (measurement, segmentation)
|
|
2. vision for perception, interpretation (labeling)
|
|
3. search and organization (retrieval, image or video archives)
|
|
|
|
### What is image
|
|
|
|
A 2d array of numbers.
|
|
|
|
### Vision is hard
|
|
|
|
connection to graphics.
|
|
|
|
computer vision need to generate the model from the image.
|
|
|
|
#### Are A and B the same color?
|
|
|
|
It depends on the context what you mean by "the same".
|
|
|
|
todo
|
|
|
|
#### Chair detector example.
|
|
|
|
double for loops.
|
|
|
|
#### Our visual system is not perfect.
|
|
|
|
Some optical illusion images.
|
|
|
|
todo, embed images here.
|
|
|
|
### Ridiculously brief history of computer vision
|
|
|
|
1960s: interpretation of synthetic worlds
|
|
1970s: some progress on interpreting selected images
|
|
1980s: ANNs come and go; shift toward geometry and increased mathematical rigor
|
|
1990s: face recognition; statistical analysis in vogue
|
|
2000s: becoming useful; significant use of machine learning; large annotated datasets available; video processing starts.
|
|
2010s: Deep learning with ConvNets
|
|
2020s: String synthesis; continued improvement across tasks, vision-language models.
|
|
|
|
## How computer vision is used now
|
|
|
|
### OCR, Optical Character Recognition
|
|
|
|
Technology to convert scanned docs to text.
|
|
|
|
|