Enabling Efficient Training of Convolutional Neural Networks for Histopathology Images

Alali, Mohammed H.; Roohi, Arman; Deogun, Jitender S.

doi:10.1007/978-3-031-13321-3_47

Cited by 1 publication

(4 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, for sparsity on RGB, the accuracy was 0.86 on <2:2>. [29]. It was shown that with a reduction of 77% of MACs, the error rate is lower by 1% from training on the original RGB color mode compared to training on sparsity on grayscale.…”

Section: Low Bit-width Precision Resultsmentioning

confidence: 96%

“…The PCam results shown in Table 5 are as reported in [29]. We performed experiments on MHIST data on the model used in [29]. The results summarized in Table 5 show that, for the PCam dataset, the accuracy increases by a maximum of 0.04 on the <2:2> configuration in the sparsity on grayscale color mode when our model is used.…”

Section: Low Bit-width Precision Resultsmentioning

confidence: 99%

“…Table 5 shows the inference results of ResNet56 when trained on low bit-width based on DoReFa-Net. In Table 5, we compare the best results obtained by [29] to our results for the same configurations. The PCam results shown in Table 5 are as reported in [29].…”

Section: Low Bit-width Precision Resultsmentioning

confidence: 99%

“…In Table 5, we compare the best results obtained by [29] to our results for the same configurations. The PCam results shown in Table 5 are as reported in [29]. We performed experiments on MHIST data on the model used in [29].…”

Section: Low Bit-width Precision Resultsmentioning

confidence: 99%

See 3 more Smart Citations

Enabling Intelligent IoTs for Histopathology Image Analysis Using Convolutional Neural Networks

et al. 2022

Self Cite

View full text Add to dashboard Cite

Medical imaging is an essential data source that has been leveraged worldwide in healthcare systems. In pathology, histopathology images are used for cancer diagnosis, whereas these images are very complex and their analyses by pathologists require large amounts of time and effort. On the other hand, although convolutional neural networks (CNNs) have produced near-human results in image processing tasks, their processing time is becoming longer and they need higher computational power. In this paper, we implement a quantized ResNet model on two histopathology image datasets to optimize the inference power consumption. We analyze classification accuracy, energy estimation, and hardware utilization metrics to evaluate our method. First, the original RGB-colored images are utilized for the training phase, and then compression methods such as channel reduction and sparsity are applied. Our results show an accuracy increase of 6% from RGB on 32-bit (baseline) to the optimized representation of sparsity on RGB with a lower bit-width, i.e., <8:8>. For energy estimation on the used CNN model, we found that the energy used in RGB color mode with 32-bit is considerably higher than the other lower bit-width and compressed color modes. Moreover, we show that lower bit-width implementations yield higher resource utilization and a lower memory bottleneck ratio. This work is suitable for inference on energy-limited devices, which are increasingly being used in the Internet of Things (IoT) systems that facilitate healthcare systems.

show abstract

Section: Low Bit-width Precision Resultsmentioning

confidence: 96%

Section: Low Bit-width Precision Resultsmentioning

confidence: 99%