PhotoTechEDU Day 30: Imaging optics for the next decade
Google Tech Talks November, 7 2007 Digital cameras in their many forms will continue to be one of the primary drivers towards new technologies in optics as well as improvements of classical technologies. This has been well illustrated in the past 5-10 years which has seen, for example, the development of compression molded glass aspheric lenses for improved performance and packaging. The incorporation of injection molded plastic lenses and possibly hybrid refractive/diffractive surfaces will grow. Furthermore, as the trend continues towards smaller pixels as well as more pixels in a given sensor, the imaging optics will be further driven towards higher image quality. Zoom lenses will increase in their zoom range, yet there will be a continuing emphasis towards smaller and smaller packaging. The optics and their associated mechanics will need to be more robust with respect to stray light such as flare, glare, ghost images, and other undesirable image anomalies. And our optics must be more robust with respect to environmental effects such as thermal soaks and gradients. And with all of the above, customers will want lower cost too. It is going to be a fun ride over the next 5-10 years so fasten your seat belt and hold on real tight to the safety bar! Speaker: Robert E. Fischer, CEO OPTICS 1, Inc. Westlake Village, CA Bob Fisher is CEO of Optics 1, a past president of the SPIE, and a winner of that society's highest award, the Gold Medal for outstanding engineering or scientific accomplishments in optics andelectro-optics. Mr. Fischer's technical interests are in optical system design and engineering, in particular lens design. He is also interested in optical component and system manufacturing, assembly, and testing. His interests extend from the deep UV through the visible and on to the thermal infrared. He is known for his tireless efforts to advance optical science, engineering and scholarship. He served as a book editor of the McGraw-Hill Series on Optical and Electro-Optical Engineering, and as executive editor of OE Reports, bringing timely and practical information to professionals in the field.
Going Back and Beyond: Emerging (Old) Threats in LLM Privacy and Poisoning
Code Health Guardian
Private Adaptations of Large Language Models
Atomic Facts to Structured Knowledge: Rethinking Unlearning & Jailbreaking in Large Language Models
The Limits and Possibilities of One Run Auditing
Local Node Differential Privacy
POPri: Private Federated Learning using Preference-Optimized Synthetic Data
Worst-Case Membership Inference of Language Models
Differentially Private Table-Image Multimodal Data Generation
Cascading Adversarial Bias from Injection to Distillation in Language Models
Privacy Amplification for Correlated-Noise Mechanisms via b-Min-Sep Subsampling
Stable Estimators for Fast Private Statistics
Threat Models for Memorization: Privacy, Copyright, and Everything In-Between
Privacy Ripple Effects from Adding or Removing Personal Information in Language Model Training
Disparate Privacy Risks from Medical AI - An Investigation into Patient-level Privacy Risk
Understanding Trade-offs Between Worst-case Differential Privacy Guarantees & Real Threat Models
Privacy Auditing of Large Language Models
Streaming Private Continual Counting via Binning
