Detection of environmental contamination such as trace-level toxic heavy metal ions mostly relies on bulky and costly analytical instruments. However, a considerable global need exists for portable, rapid, specific, sensitive, and cost-effective detection techniques that can be used in resource-limited and field settings. Here we introduce a smart-phone-based hand-held platform that allows the quantification of mercury(II) ions in water samples with parts per billion (ppb) level of sensitivity. For this task, we created an integrated opto-mechanical attachment to the built-in camera module of a smart-phone to digitally quantify mercury concentration using a plasmonic gold nanoparticle (Au NP) and aptamer based colorimetric transmission assay that is implemented in disposable test tubes. With this smart-phone attachment that weighs <40 g, we quantified mercury(II) ion concentration in water samples by using a two-color ratiometric method employing light-emitting diodes (LEDs) at 523 and 625 nm, where a custom-developed smart application was utilized to process each acquired transmission image on the same phone to achieve a limit of detection of ∼3.5 ppb. Using this smart-phone-based detection platform, we generated a mercury contamination map by measuring water samples at over 50 locations in California (USA), taken from city tap water sources, rivers, lakes, and beaches. With its cost-effective design, field-portability, and wireless data connectivity, this sensitive and specific heavy metal detection platform running on cellphones could be rather useful for distributed sensing, tracking, and sharing of water contamination information as a function of both space and time.
DNA imaging techniques using optical microscopy have found numerous applications in biology, chemistry and physics and are based on relatively expensive, bulky and complicated set-ups that limit their use to advanced laboratory settings. Here we demonstrate imaging and length quantification of single molecule DNA strands using a compact, lightweight and cost-effective fluorescence microscope installed on a mobile phone. In addition to an optomechanical attachment that creates a high contrast dark-field imaging setup using an external lens, thin-film interference filters, a miniature dovetail stage and a laser-diode for oblique-angle excitation, we also created a computational framework and a mobile phone application connected to a server back-end for measurement of the lengths of individual DNA molecules that are labeled and stretched using disposable chips. Using this mobile phone platform, we imaged single DNA molecules of various lengths to demonstrate a sizing accuracy of <1 kilobase-pairs (kbp) for 10 kbp and longer DNA samples imaged over a field-of-view of ∼2 mm2.
There is an increasing need to bring machine learning to a wide diversity of hardware devices. Current frameworks rely on vendor-specific operator libraries and optimize for a narrow range of server-class GPUs. Deploying workloads to new platforms -such as mobile phones, embedded devices, and accelerators (e.g., FPGAs, ASICs) -requires significant manual effort. We propose TVM, a compiler that exposes graph-level and operator-level optimizations to provide performance portability to deep learning workloads across diverse hardware back-ends. TVM solves optimization challenges specific to deep learning, such as high-level operator fusion, mapping to arbitrary hardware primitives, and memory latency hiding. It also automates optimization of low-level programs to hardware characteristics by employing a novel, learning-based cost modeling method for rapid exploration of code optimizations. Experimental results show that TVM delivers performance across hardware back-ends that are competitive with state-ofthe-art, hand-tuned libraries for low-power CPU, mobile GPU, and server-class GPUs. We also demonstrate TVM's ability to target new accelerator back-ends, such as the FPGA-based generic deep learning accelerator.The system is open sourced and in production use inside several major companies.
Specialized Deep Learning (DL) acceleration stacks, designed for a specific set of frameworks, model architectures, operators, and data types, offer the allure of high performance while sacrificing flexibility. Changes in algorithms, models, operators, or numerical systems threaten the viability of specialized hardware accelerators.We propose VTA, a programmable deep learning architecture template designed to be extensible in the face of evolving workloads. VTA achieves this flexibility via a parametrizable architecture, two-level ISA, and a JIT compiler. The two-level ISA is based on (1) a task-ISA that explicitly orchestrates concurrent compute and memory tasks and (2) a microcode-ISA which implements a wide variety of operators with single-cycle tensor-tensor operations. Next, we propose a runtime system equipped with a JIT compiler for flexible code-generation and heterogeneous execution that enables effective use of the VTA architecture.VTA is integrated and open-sourced into Apache TVM, a state-ofthe-art deep learning compilation stack that provides flexibility for diverse models and divergent hardware backends. We propose a flow that performs design space exploration to generate a customized hardware architecture and software operator library that can be leveraged by mainstream learning frameworks. We demonstrate our approach by deploying optimized deep learning models used for object classification and style transfer on edge-class FPGAs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.