Abstract
Approximation approaches have been developed for very large-scale integration architecture to limit power consumption and increase effective throughput. The full adder (FA) and block-based designs are the approximate adder designs widely used in the existing techniques to accelerate the additions by partitioning a long carry propagation chain. However, these techniques require more power and area than the standard FA-based techniques. To overcome this issue, this paper presents an error-reduced carry prediction approximate full adder (ERCPAA) technique to achieve faster additions and performance by adding a constant truncation mechanism with an error reduction strategy. Deep learning and neural network technology had made great progress in recent years for field programmable gate array implementation and real-time object detection. It is essential to maintain high accuracy rates in image processing applications, as any errors introduced during the processing can significantly impact the quality of the output image and impact the hardware implementation. Therefore, we propose an ERCPAA that utilizes a lightweight convolutional neural network (ERCPAA–lightweight CNN) to fit well in a fixed-point CNN accelerator architecture. A lightweight CNN that is both thin (with fewer feature maps per layer) and deep (four layers) with just one fully connected hidden layer can enable faster training while achieving higher accuracy. In addition, in a few positions of the higher-order bits of the incorrect part, a full adder (FA) cell is used to simplify the traditional one-bit FA cell, creating an approximate summation and carry. The proposed method is verified by randomly selecting specific images from the Fruit 360, ImageNet, and the Caltech 256 Image Dataset. The proposed model offers an normalized mean error distance, error rate, and mean relative error distance values of 0.154, 6.55%, and 4.6E-4. The recognition accuracy of the proposed model in the ImageNet database is 96% with an execution time of 1420 µs.
Similar content being viewed by others
Data availability
The data that supports the findings of this study is available from the corresponding author upon reasonable request.
References
Priyadharshni M, Raj Gupta A, Nithish Kumar V, Kumaravel S (2021) An error-efficient and low-complexity approximate multi-bit adder for image processing applications. Int J Circuit Theory Appl 49(8):2373–2381
Padmini GR, Rajesh O, Raghu K, Sree NM, Apurva C (2021) Design and analysis of 8-bit ripple carry adder using nine transistor full adder. In: 2021 7th international conference on advanced computing and communication systems (ICACCS), vol 1. IEEE, pp 1982–1987
Reshadi ZZKNM, Keshavarziyan P (2021) Efficient symmetrical imprecise 1-bit full adder cells using CNFET. Technology for Image Processing Applications
Cui X, Liu W, Wang S, Swartzlander EE, Lombardi F (2018) Design of high-speed wide-word hybrid parallel-prefix/carry-select and skip adders. J Signal Process Syst 90(3):409–419
Kadam DB, Pandyaji KK, Liyakat KKS (2022) Implementation of carry select adder (CSLA) for area, delay and power minimization. Telematique 21:5461–5474
Thakur G, Sohal H, Jain S (2021) A novel ASIC-based variable latency speculative parallel prefix adder for image processing application. Circuits Syst Signal Process 40(11):5682–5704
Palanisamy G, Krishnasamy Natarajan V, Sundaram K (2019) Area-efficient parallel adder with faithful approximation for image and signal processing applications. IET Image Proc 13(13):2587–2594
Nishok VS, Poongodi P, Vijeyakumar KN (2018) Design and performance estimation of efficient approximate carry select adder. Appl Inf Sci Int J 12(6):1219–1225
Kowsalya T (2020) Area and power efficient pipelined hybrid merged adders for customized deep learning framework for FPGA implementation. Microprocess Microsyst 72:102906
Jothin R, Mohamed MP, Vasanthanayaki C (2020) High performance compact energy efficient error tolerant adders and multipliers for 16-bit image processing applications. Microprocess Microsyst 78:103237
Nagarajan M, Muthaiah R, Teekaraman Y, Kuppusamy R, Radhakrishnan A (2022) Power and Area Efficient Cascaded Effectless GDI Approximate Adder for Accelerating Multimedia Applications Using Deep Learning Model. Computational Intelligence and Neuroscience 2022.
Lakshmi TSS (2021) Speculative carry addition performance improvement and area optimization using modified carry generators. CVR J Science Technol 21(1):55–61
Wang Y, Pai C, Song X (2002) The design of hybrid carry-lookahead/carry-select adders. IEEE Trans Circuits Syst II Analog Digit Signal Process 49(1):16–24
Khaksari O, Akbari Ebrahimi B (2022) BEAD: bounded error approximate adder with carry and sum speculations. Integration 88:353–361
Javadi MHS, Yalame MH, Mahdiani HR (2020) Small constant mean-error imprecise adder/multiplier for efficient VLSI implementation of MAC-based applications. IEEE Trans Comput 69(9):1376–1387
Mendez T, Nayak SG, Kumar PV, Kedlaya KV (2022) Performance metric evaluation of error-tolerant adders for 2D image blending. Electronics 11(15):2461
Qiao GC, Hu SG, Chen TP, Rong LM, Ning N, Yu Q, Liu Y (2020) STBNN: hardware-friendly spatio-temporal binary neural network with high pattern recognition accuracy. Neurocomputing 409:351–360
Mani VRS, Saravanaselvan A, Arumugam N (2022) Performance comparison of CNN, QNN and BNN deep neural networks for real-time object detection using ZYNQ FPGA node. Microelectron J 119:105319
Quan H, Dong J, Zhan C, Wang Q, He Y, Qian X (2022) Lightweight CNN based on non-rigid SFM for 3D reconstruction of medical images. In: 2022 IEEE international conference on bioinformatics and biomedicine (BIBM), IEEE, pp 791–794
Li Y, Zhang S, Wang WQ (2020) A lightweight faster R-CNN for ship detection in SAR images. IEEE Geosci Remote Sens Lett 19:1–5
Pomanysochka Y, Kondratenko Y, Kondratenko G, Sidenko I (2019) Soft computing techniques for noise filtration in the image recognition processes. In: 2019 IEEE 2nd Ukraine conference on electrical and computer engineering (UKRCON), IEEE, pp 1189–1195
Kondratenko Y, Sidenko I, Kondratenko G, Petrovych V, Taranov M, Sova I (2020) Artificial neural networks for recognition of brain tumors on MRI images. In: international conference on information and communication technologies in education, research, and industrial applications, Springer International Publishing, Cham, pp 119-140
Sheremet A, Kondratenko Y, Sidenko I, Kondratenko G (2021) Diagnosis of lung disease based on medical images using artificial neural networks. In: 2021 IEEE 3rd Ukraine conference on electrical and computer engineering (UKRCON), IEEE, pp 561–565
Wang W (2023) Mean-square exponential input-to-state stability of stochastic fuzzy delayed Cohen-Grossberg neural networks. J Exp Theor Artif Intell. https://doi.org/10.1080/0952813X.2023.2165725
Liu B (2017) Finite-time stability of CNNs with neutral proportional delays and time-varying leakage delays. Math Methods Appl Sci 40(1):167–174
Lee J, Seo H, Seok H, Kim Y (2021) A novel approximate adder design using error reduced carry prediction and constant truncation. IEEE Access 9:119939–119953
Mahdiani HR, Ahmadi A, Fakhraie SM, Lucas C (2009) Bio-inspired imprecise computational blocks for efficient VLSI implementation of soft-computing applications. IEEE Trans Circuits Syst I Regul Pap 57(4):850–862
Dalloo A, Najafi A, Garcia-Ortiz A (2018) Systematic design of an approximate adder: the optimized lower part constant-OR adder. IEEE Trans Very Large Scale Integr (VLSI) Syst 26(8):1595–1599
Mazahir S, Ayub MK, Hasan O, Shafique M (2019) Probabilistic error analysis of approximate adders and multipliers. In: Approximate circuits: methodologies and CAD, pp 99–120
Haque WA, Arefin S, Shihavuddin ASM, Hasan MA (2021) DeepThin: a novel lightweight CNN architecture for traffic sign recognition without GPU requirements. Expert Syst Appl 168:114481
Oltean M (2021) Fruits 360. Kaggle. https://www.kaggle.com/datasets/moltean/fruits. Accessed 25 Nov 2022
Li J (2018) Caltech 256 image dataset. Kaggle. https://www.kaggle.com/datasets/jessicali9530/caltech256. Accessed 25 Nov 2022
Funding
Not applicable.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Human and animal rights
This article does not contain any studies with human or animal subjects performed by any of the authors.
Informed consent
Informed consent was obtained from all individual participants included in the study.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Nishanth, R., Sulochana, C.H. A novel lightweight CNN-based error-reduced carry prediction approximate full adder design for multimedia applications. Neural Comput & Applic 36, 6421–6440 (2024). https://doi.org/10.1007/s00521-023-09316-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-023-09316-z