Real time vision by FPGA implemented CNNs

Lopez-Garcia, J.C.; Moreno-Armendáriz, Marco A.; Riera-Babures, J.; Balsi, Marco; Vilasís-Cardona, X.

doi:10.1109/ecctd.2005.1522965

Cited by 4 publications

(3 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nowadays, we are involved in a project of a new robot in which CNNs are emulated on an FPGA, and this should allow a faster processing of the image. For example, in [19] is shown that an FPGA can emulate CNNs up to 9 times faster than a DSP can do (of course, these data depend on the characteristics of the two devices). Anyway, such projects should enhance the development of dedicated CNN hardware as well as spark the search for CNN solutions to a number of problems.…”

Section: Discussionmentioning

confidence: 98%

Robot vision with cellular neural networks: a practical implementation of new algorithms

Pazienza

Ponce-García

Balsi

et al. 2006

Circuit Theory & Apps

View full text Add to dashboard Cite

SUMMARYCellular neural networks (CNNs) are well suited for image processing due to the possibility of a parallel computation. In this paper, we present two algorithms for tracking and obstacle avoidance using CNNs. Furthermore, we show the implementation of an autonomous robot guided using only real-time visual feedback; the image processing is performed entirely by a CNN system embedded in a digital signal processor (DSP). We successfully tested the two algorithms on this robot.

show abstract

Section: Discussionmentioning

confidence: 98%

Robot vision with cellular neural networks: a practical implementation of new algorithms

Pazienza

Ponce-García

Balsi

et al. 2006

Circuit Theory & Apps

View full text Add to dashboard Cite

show abstract

“…As far as our project is concerned, we first focused on the CNN-UM emulation on a Xilinx FPGA programmed in Handel-C [8]; then, we moved to a Texas Instruments DSP programmed in C [9], whose performances were better than the previous system but far from being considered real-time processing; finally, we returned to a FPGA but this time programmed in VHDL [10], obtaining the best results so far.…”

Section: State Of the Art Of Cnn-um Emulation On Fpgamentioning

confidence: 99%

“…In particular, we have considered a Xilinx Spartan II 200K FPGA programmed in Handel-C [8], a TMS320c32 DSP programmed in C [9], and the non-optimized version of the FPGA programmed in VHDL [10]. The results are summarized in Fig.…”

Section: B Comparison With Previous Approachesmentioning

confidence: 99%

Optimized cellular neural network universal machine emulation on FPGA

Pazienza

Bellana-Camanes

Riera-Babures

et al. 2007

2007 18th European Conference on Circuit Theory and Design

View full text Add to dashboard Cite

An FPGA architecture to emulate a single-layer Cellular Neural Network -Universal Machine (CNN-UM) is proposed. It is based on a fast realization of the CNN convolution operation on the parallel hardware of the FPGA. The setup is capable of performing a CNN iteration over a 30×30 pixel image in less than 30 µs. Moreover, this platform has been used to realize the visual system of an autonomous mobile robot.

show abstract