-
TabConv: Low-Computation CNN Inference via Table Lookups
Authors:
Neelesh Gupta,
Narayanan Kannan,
Pengmiao Zhang,
Viktor Prasanna
Abstract:
Convolutional Neural Networks (CNNs) have demonstrated remarkable ability throughout the field of computer vision. However, CNN inference requires a large number of arithmetic operations, making them expensive to deploy in hardware. Current approaches alleviate this issue by develo** hardware-supported, algorithmic processes to simplify spatial convolution functions. However, these methods still…
▽ More
Convolutional Neural Networks (CNNs) have demonstrated remarkable ability throughout the field of computer vision. However, CNN inference requires a large number of arithmetic operations, making them expensive to deploy in hardware. Current approaches alleviate this issue by develo** hardware-supported, algorithmic processes to simplify spatial convolution functions. However, these methods still heavily rely on matrix multiplication, leading to significant computational overhead. To bridge the gap between hardware, algorithmic acceleration, and approximate matrix multiplication, we propose TabConv, a novel, table-based approximation for convolution to significantly reduce arithmetic operations during inference. Additionally, we introduce a priority masking technique based on cosine similarity to select layers for table-based approximation, thereby maintaining the model performance. We evaluate our approach on popular CNNs: ResNet-18, ResNet-34, and NetworkInNetwork (NIN). TabConv preserves over 93% of the original model's performance while reducing arithmetic operations by 36.5%, 25.8%, and 99.4% for ResNet-18 on CIFAR-10, CIFAR-100, and MNIST, respectively, 35.6% and 99.3% for ResNet-34 on CIFAR-10 and MNIST, and 98.9% for NIN on MNIST, achieving low-computation inference.
△ Less
Submitted 8 April, 2024;
originally announced April 2024.
-
XAI meets Biology: A Comprehensive Review of Explainable AI in Bioinformatics Applications
Authors:
Zhongliang Zhou,
Mengxuan Hu,
Mariah Salcedo,
Nathan Gravel,
Wayland Yeung,
Aarya Venkat,
Dongliang Guo,
Jielu Zhang,
Natarajan Kannan,
Sheng Li
Abstract:
Artificial intelligence (AI), particularly machine learning and deep learning models, has significantly impacted bioinformatics research by offering powerful tools for analyzing complex biological data. However, the lack of interpretability and transparency of these models presents challenges in leveraging these models for deeper biological insights and for generating testable hypotheses. Explaina…
▽ More
Artificial intelligence (AI), particularly machine learning and deep learning models, has significantly impacted bioinformatics research by offering powerful tools for analyzing complex biological data. However, the lack of interpretability and transparency of these models presents challenges in leveraging these models for deeper biological insights and for generating testable hypotheses. Explainable AI (XAI) has emerged as a promising solution to enhance the transparency and interpretability of AI models in bioinformatics. This review provides a comprehensive analysis of various XAI techniques and their applications across various bioinformatics domains including DNA, RNA, and protein sequence analysis, structural analysis, gene expression and genome analysis, and bioimaging analysis. We introduce the most pertinent machine learning and XAI methods, then discuss their diverse applications and address the current limitations of available XAI tools. By offering insights into XAI's potential and challenges, this review aims to facilitate its practical implementation in bioinformatics research and help researchers navigate the landscape of XAI tools.
△ Less
Submitted 10 December, 2023;
originally announced December 2023.
-
A Theoretical Analysis of Granulometry-based Roughness Measures on Cartosat DEMs
Authors:
Nagajothi Kannan,
Sravan Danda,
Aditya Challa,
Daya Sagar B S
Abstract:
The study of water bodies such as rivers is an important problem in the remote sensing community. A meaningful set of quantitative features reflecting the geophysical properties help us better understand the formation and evolution of rivers. Typically, river sub-basins are analysed using Cartosat Digital Elevation Models (DEMs), obtained at regular time epochs. One of the useful geophysical featu…
▽ More
The study of water bodies such as rivers is an important problem in the remote sensing community. A meaningful set of quantitative features reflecting the geophysical properties help us better understand the formation and evolution of rivers. Typically, river sub-basins are analysed using Cartosat Digital Elevation Models (DEMs), obtained at regular time epochs. One of the useful geophysical features of a river sub-basin is that of a roughness measure on DEMs. However, to the best of our knowledge, there is not much literature available on theoretical analysis of roughness measures. In this article, we revisit the roughness measure on DEM data adapted from multiscale granulometries in mathematical morphology, namely multiscale directional granulometric index (MDGI). This measure was classically used to obtain shape-size analysis in greyscale images. In earlier works, MDGIs were introduced to capture the characteristic surficial roughness of a river sub-basin along specific directions. Also, MDGIs can be efficiently computed and are known to be useful features for classification of river sub-basins. In this article, we provide a theoretical analysis of a MDGI. In particular, we characterize non-trivial sufficient conditions on the structure of DEMs under which MDGIs are invariant. These properties are illustrated with some fictitious DEMs. We also provide connections to a discrete derivative of volume of a DEM. Based on these connections, we provide intuition as to why a MDGI is considered a roughness measure. Further, we experimentally illustrate on Lower-Indus, Wardha, and Barmer river sub-basins that the proposed features capture the characteristics of the river sub-basin.
△ Less
Submitted 16 July, 2021;
originally announced July 2021.
