A specification-by-example toolkit that generates formal network specifications using only input-output examples.

Published: 11/19/2024

A physical embodiment of a locally-learning neural network, which learns faster than current CPU-based nets.

Published: 10/24/2024

A radio frequency-based 3D imaging system that can generate LiDAR-comparable images with reduced power and processing needs.

Published: 8/26/2024

SCALPEL is a lightweight optimization tool for automatically compartmentalizing policies for hardware-accelerated enforcement in a tagged architecture. The tool also creates a layer or protection (or hardening) by learning and allowing certain privileges based on learned expectations.

Published: 8/2/2024

The collective genomes of the human gut microbiome, known as the metagenome, encode peptides that are effective against antibiotic-resistant bacteria.

Published: 6/26/2024

The course of radiation treatment of cancer patients has three major phases:

1. Diagnostic and Prescription;
2. Simulation and Quality Assurance; and
3. Delivery.

In Simulation and Quality Assurance (QA), a specific plan on how to deliver the prescribed radiation to the tumor is developed. Penn scientists have developed a software-based virtual Intensity Modulated Radiation Therapy (IMRT) Quality Assurance model that extracts features associated with failure modes (between the treatment planning system and the linear accelerators) and uses machine learning to learn from pass plans in order to accurately predict the gamma passing rates. This method could save on the time and cost of QA, and reduce the number of replans needed.

Published: 6/24/2024

Novel computational architecture designs to reduce the latency time to process large volumes of data utilizing the reconfiguration of memory and storage; streamlining read/write functions to include computational logic within the register file; and programmable schedule and memory utilization within a configurable load/store unit.

Published: 11/8/2024

A flexible photonics processor enabling on-chip optimization of light-speed information processing without requiring complex fabrication.

Published: 5/7/2024

A versatile system that optimizes and adapts data shuffling in distributed systems, simplifying complex operations and enhancing overall performance.

Published: 6/18/2024

AAR-DL (Automatic Anatomy Recognition – Deep Learning) is a hybrid intelligence recognition and delineation segmentation model that outputs automatically delineated organs in medical images by combining natural intelligence’s (i.e., human knowledge’s) and artificial intelligence's strengths to map organs and their boundaries in the given images.

Published: 6/24/2024