"Back in my Day..."
Name 2 applications of computer vision that are used in every day life.
Sample responses: Smart checkouts, self driving cars, doorbells detecting people
Computer Vision originated in which century?
a. 18th Century
b. 19th Century
c. 20th Century
d. 21st Century
c. 20th Century
Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton are known for creating which groundbreaking CV model?
AlexNet
The Goal of Computer vision is to extract _______ or __________ from _______
meaning / information / images
Name one part in the process of creating/using ML that could potentially introduce biases
Data sets used
Generative AI in healthcare can be used to synthesize ________ for training models without violating patient privacy.
Fake data / Medical Images
Early computer vision systems in the 1960s used this approach instead of machine learning.
a. Handcrafted rules / Classical algorithms
b. Convolutional neural networks
c. Generative models
d. Reinforcement learning
a. Handcrafted rules / Classical algorithms
Fei-Fei Li and colleagues at Princeton/Stanford created the ImageNet project, which fueled modern computer vision breakthroughs. It was launched in ________.
2009
Name at least 2 CV techniques
Sample Responses: Image classification, Object detection, Semantic segmentation, Contextual image analysis
The viral bunnies on a trampoline video is an example of what type of concern regarding AI generated images?
https://www.reddit.com/r/Bunnies/comments/1mjvetv/a_viral_security_cam_clip_of_rabbits_bouncing_on/
Misinformation
In autonomous vehicles, LiDAR is often combined with computer vision to improve ________.
Depth Perception
The shift from handcrafted features to deep learning was accelerated by which dataset benchmark?
a. COCO
b. ImageNet
c. CIFAR-10
b. ImageNet
This AI researcher contributed to the “Attention Is All You Need” paper that introduced the transformer architecture.
a. Geoffrey Hinton
b. Yann LeCun
c. Ian Goodfellow
d. Ashish Vaswani
d. Ashish Vaswani
Typically, in images and image processing for a computer, an image is an array of numeric pixel values. Each pixel has a value between ____ and ____
0 (Black), 255(Black)
Note: This is typical, but not always the case. Some applications use "deep" images that use more bits per pixel.
Switzerland has adopted the European Union's Entry/Exit System (EES). This system replaces manual passport stamps with biometric facial scans and fingerprinting. Name 2 possible ethical concerns that come with this new development.
Sample Responses: Risk this data could be used for broader surveillance beyond border control / Bias in facial recognition systems (potential discrimination) / Accountability issues
Optical character recognition (OCR) is a common application of computer vision for which type of data?
a. Text / Documents
b. Satellite imagery
c. Facial expressions
a. Text / Documents
The 1998 LeNet-5 model was designed primarily for which type of image recognition task?
a. Traffic sign recognition
b. Handwritten digit recognition (MNIST)
c. Facial recognition
d. Object detection
b. Handwritten digit recognition (MNIST)
Fei-Fei Li is widely recognized for her contributions to which area in computer vision?
a. Reinforcement learning
b. Optical character recognition
c. ImageNet dataset and large-scale image classification
d. GAN-based image generation
c. ImageNet dataset and large-scale image classification
The training process repeats over multiple _______ until an optimal set of weights has been learned, then the weights are saved and can be used to predict labels for images
epochs
True or False? In the EU if a company uses an AI voice clone of a celebrity to narrate an advertisement, they do not have to state that the audio is synthetic unless it is accompanied by video.
False!
Note: The EU AI Act imposes a binding transparency requirement (Article 50), requiring deployers(users) of AI systems that generate deepfakes to disclose (label) that the content has been artificially generated or manipulated. And the fines for serious violations can reach up to 6% of a company’s global turnover.
BJ’s autonomous robot(s), Tally, monitor inventory during live store operations. Tally uses computer vision to identify empty or low-stock shelves, detect misplaced products, and provide real-time inventory updates. Name 1
Sample Answers: object detection, depth estimation, and shelf segmentation
The “AI winter” in the 1980s and 1990s was caused partly by the limitations of which model?
a. Convolutional Neural Network (CNN)
b. Perceptron / Shallow Neural Networks
c. Recurrent Neural Network (RNN)
d. Generative Adversarial Network (GAN)
b. Perceptron / Shallow Neural Networks
Geoffrey Hinton is often called the “godfather of ________.”
a. Deep learning
b. Reinforcement learning
c. Computer vision
d. Natural language processing
a. Deep learning
The key differences between Embedded values in vision transformers vs Natural Language Processing are based on features like
a. color & shape
b. contrast & texture
c. None of the above
d. All of the above
d. All of the above
Context: Deepfakes are a misuse risk of computer vision and generative AI involving fake images and videos.
True or false? The TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act) was enacted on May 19, 2025, after President Trump signed it into law. The Act criminalizes the knowing publication or threat to publish NCII, including both authentic and AI-generated deepfakes.
True!
Note: This bill generally prohibits the nonconsensual online publication of intimate visual depictions of individuals, both authentic and computer-generated, and requires certain online platforms to promptly remove such depictions upon receiving notice of their existence.