Knowledge Vault 5 /82 - CVPR 2023
History and Future of Artificial Intelligence and Computer Vision
Linda Shapiro, Jamie Shotton, Chelsea Finn, Daniel Huttenlocher
< Resume Image >

Concept Graph & Resume using Claude 3 Opus | Chat GPT4o | Llama 3:

graph LR classDef people fill:#f9d4d4, font-weight:bold, font-size:14px classDef history fill:#d4f9d4, font-weight:bold, font-size:14px classDef challenges fill:#d4d4f9, font-weight:bold, font-size:14px classDef advice fill:#f9f9d4, font-weight:bold, font-size:14px classDef future fill:#f9d4f9, font-weight:bold, font-size:14px A[History and Future
of Artificial Intelligence
and Computer Vision] --> B[Hottenlacher: Cornell, MIT prof,
computer vision pioneer 1] A --> C[Shapiro: UW prof, segmentation,
matching, robotics, medical 2] A --> D[Finn: Stanford prof, robot
learning through interaction 3] A --> E[Shotton: WAVE Chief Scientist,
pose estimation, 3D reconstruction 4] A --> F[Computer vision history:
pattern recognition to CVPR 5] A --> G[Scaling challenges: submissions,
reviews, regional conferences 6] A --> H[Preserving history: teaching,
oral histories, personalities 7] A --> I[Revisiting old methods:
random forests, supervised limits 8] A --> J[Analogies: graphics, pharma,
aerospace, exhibits growth 9] A --> K[Future: household robots,
uncertainty, embodied AI 10] A --> L[Academia resource constraints,
collaboration opportunities 11] A --> M[Language importance in
complex scene understanding 12] A --> N[AI systems concerns:
regulation, fake references 13] A --> O[Too many papers:
competitiveness discourages ideas 14] A --> P[Advice: motivating problems,
mentors, strengths, be different 15] class B,C,D,E people class F,G,H history class I,J challenges class K,L,M,N future class O,P advice


1.- Dan Hottenlacher: Professor at Cornell and MIT, influential in computer vision field. Now Dean of Schwarzman College of Computing at MIT.

2.- Linda Shapiro: Professor at University of Washington, made key contributions to image segmentation, matching, robot vision, medical applications.

3.- Chelsea Finn: Assistant Professor at Stanford, focuses on methods for robots and agents to develop intelligence through learning and interaction.

4.- Jamie Shotton: Chief Scientist at WAVE, worked on pose estimation and 3D reconstruction from Kinect depth images at Microsoft.

5.- History of computer vision: Pattern recognition conference to CVPR. Field has evolved with machine learning's increasing influence.

6.- Scaling challenges: Growing number of submissions, review process difficulties, need for more regional conferences, sustainable scaling.

7.- Preserving history: Importance of teaching historical works, capturing oral histories, conveying researcher personalities through their work.

8.- Revisiting old methods: Potential value in efficiency of random forests, limitations of purely supervised learning frameworks.

9.- Analogy to other fields: Similar to graphics, pharmaceutical industry, aerospace. Transition from small community to large with exhibits.

10.- Future challenges and applications: Robotics (household humanoid robots), uncertainty quantification, task specification, open-world embodied AI.

11.- Resource constraints in academia: Need for industry, government, academia collaboration on shared infrastructure. Innovation opportunities under constraints.

12.- Importance of language in computer vision: Valuable for complex scene understanding, but less critical for low-level perception and motor skills.

13.- Concerns about AI systems and values: Risk of over and under regulation due to lack of understanding. Fake references and facts from LLMs.

14.- Too many papers submitted to conferences: Due to competitiveness and importance of top conferences. Limits may discourage new ideas and young researchers.

15.- Advice for aspiring researchers: Find motivating problems, seek good mentors and collaborators, leverage strengths while improving weaknesses, be different.

Knowledge Vault built byDavid Vivancos 2024