Computational Imaging and Artificial Intelligence: The Next Revolution of Mobile Vision

Suo, Jinli; Zhang, Weihang; Gong, Jie; Yuan, Xin; Brady, David J.; Dai, Qionghai

doi:10.48550/arxiv.2109.08880

Cited by 3 publications

(3 citation statements)

References 295 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Today, combining 5G technology, computational power, and new methods to develop image recognition software allowed cameras to recognize and count objects, map environments, and help robots and autonomous vehicles with navigation, transportation, and tracking. (2) In the 1960s, researchers were extremely optimistic about the possibilities of artificial intelligence. Researchers received both private and public funds to develop and create AI.…”

Section: Cameras and Computer Visionmentioning

confidence: 99%

hopkins Studies in History of Artificial Intelligence: Perception and Sensation

Muhammed¹

2022

Preprint

View full text Add to dashboard Cite

Section: Cameras and Computer Visionmentioning

confidence: 99%

hopkins Studies in History of Artificial Intelligence: Perception and Sensation

Muhammed¹

2022

Preprint

View full text Add to dashboard Cite

“…Recently, with the rapid development of computational imaging, researchers begin to focus not only on the front-end optical components, but also on the back-end imaging recovery algorithm [1][2][3] . Computational imaging is to employ the front-end optical elements to encode the light field, and then leverage the algorithm to decode the blurry image to obtain a better result than that of directly imaging 4,5 .…”

Section: Introductionmentioning

confidence: 99%

Inverse design of metasurfaces for end-to-end computational imaging

Zhang¹,

Yu²,

Liu³

et al. 2023

Optoelectronic Imaging and Multimedia Technology IX

View full text Add to dashboard Cite

Metasurfaces, composed of two-dimensional arrays of subwavelength optical scatterers, are regarded as powerful substitutes to conventional diffractive and refractive optics. In addition, metasurfaces with powerful wavefront manipulation capabilities can steer the phase, amplitude, and polarization of light, which provides the potential to joint optimization with algorithms by encoding and decoding the light fields. In this paper, we propose an end-to-end computational imaging system which is joint optimized of metaoptics and neural networks based on the designed initial phase. We construct the forward model of the unit cell to the optical response and the inverse mapping of the optical response to the unit cell for the differentiable front-end metaoptics. Based on the appropriate initial phase, the calculation of the framework would converge faster, and the proposed system will promote the further development of metaoptics and computational imaging.

show abstract

“…The AI hardware (Neural Processing Unit (NPU)) market was valued at 8 billion USD in 2020 and is expected to grow to 84 billion USD by 2028 [1] (CAGR of 34.15%). Hardware-based AI chips are expected to play a major role in telecommunications [8], mobile vision [34], edge computing [7], augmented/virtual reality [2], health care [17], and autonomous driving [41]. Similar to code for general-purpose processors, ML models incorporate a lot of high-value intellectual property (IP) that takes a lot of time and effort to collate and develop.…”

Section: Introductionmentioning

confidence: 99%

Seculator: A Fast and Secure Neural Processing Unit

Shrivastava¹,

Sarangi²

2022

Preprint

View full text Add to dashboard Cite

Securing deep neural networks (DNNs) is a problem of significant interest since an ML model incorporates high-quality intellectual property, features of data sets painstakingly collated by mechanical turks, and novel methods of training on large cluster computers. Sadly, attacks to extract model parameters are on the rise, and thus designers are being forced to create architectures for securing such models. State-of-the-art proposals in this field take the deterministic memory access patterns of such networks into cognizance (albeit partially), group a set of memory blocks into a tile, and maintain state at the level of tiles (to reduce storage space). For providing integrity guarantees (tamper avoidance), they don't propose any significant optimizations, and still maintain block-level state.We observe that it is possible to exploit the deterministic memory access patterns of DNNs even further, and maintain state information for only the current tile and current layer, which may comprise a large number of tiles. This reduces the storage space, reduces the number of memory accesses, increases performance, and simplifies the design without sacrificing any security guarantees. The key techniques in our proposed accelerator architecture, Seculator, are to encode memory access patterns to create a small HW-based tile version number generator for a given layer, and to store layer-level MACs. We completely eliminate the need for having a MAC cache and a tile version number store (as used in related work). We show that using intelligently-designed mathematical operations, these structures are not required. By reducing such overheads, we show a speedup of 16% over the closest competing work.

show abstract

Computational Imaging and Artificial Intelligence: The Next Revolution of Mobile Vision

Cited by 3 publications

References 295 publications

hopkins Studies in History of Artificial Intelligence: Perception and Sensation

hopkins Studies in History of Artificial Intelligence: Perception and Sensation

Inverse design of metasurfaces for end-to-end computational imaging

Seculator: A Fast and Secure Neural Processing Unit

Contact Info

Product

Resources

About