Novel Arithmetics in Deep Neural Networks Signal Processing for Autonomous Driving: Challenges and Opportunities

Abstract

This paper focuses on trends, opportunities and challenges of novel arithmetics for DNN signal processing, with particular reference to assisted and autonomous driving applications. Due to strict constrains in terms of latency, dependability and security of autonomous driving, machine perception (i.e. detection or decisions tasks) based on DNN can not be implemented relying on a remote cloud access. These tasks must be performed in real-time on embedded systems on-board the vehicle, particularly for the inference phase (considering the use of DNNs pre-trained during an off-line step). When developing a DNN computing platform, the choice of the computing arithmetics matters. Moreover, functional safe applications like autonomous driving pose severe constraints on the effect that signal processing accuracy has on final rate of wrong detection/decisions. Hence, after reviewing the different choices and trade-off concerning arithmetics, both in academia and industry, we highlight the issues in implementing DNN accelerators to achieve accurate and low-complex processing of automotive sensor signals (the latter coming from diverse sources like cameras, radars, lidars, ultrasonics). The focus is on both on general-purpose operations massively used in DNN like multiply, accumulation, compare, or on specific functions like for example sigmoid or hyperbolic tangent, used for neuron activation.

Vectorizing posit operations on RISC-V for faster deep neural networks: experiments and comparison with ARM SVE

“…The application of posit numbers to Deep Neural Networks (DNN) has been independently proven to this authors and others, to perform as good as float numbers [15,25] with half the bits (or even less), as reported in…”

“…• develop our posit software library (cppPosit, see Sect. 4); • implement fast approximated activation functions, only possible when using posits, which exploit the ALU (no PPU necessary for it): [13,15]; • fast matrix-vector multiplication, thanks to vectorization (demonstrated on ARM CPUs, using SVE: [14]).…”

“…The solution proposed in this work is of interests for cases ii), iii) and iv). Indeed, In case ii) the use of ALUemulated Posit-DNNs is particularly interesting and justified (see [13,15]). Also in case iii), the approach proposed here is a viable and appealing solution, provided that the CPU supports vectorization.…”

The Italian research on HPC key technologies across EuroHPC

“…CppPosits. CppPosits [13] is a library developed at DII-University of Pisa, supporting mixed-precision and Posits arithmetic, and compliant with LLVM compiler, to increase the efficiency of AI and video computing kernels. CppPosits has been ported on ARM SVE and RISC-V with Vector extension ISA, proving that the same accuracy of FP32 can be achieved by reducing by a factor of 4 the data transfer and storage cost.…”

Photonic Integrated Reconfigurable Linear Processors as Neural Network Accelerators

“…Moreover, there are proposals to use a completely different representation for real numbers, such as the posit format introduced in 2017 [21]. Although the posit format is promising for low-precision DNNs [22][23][24], the lack of hardware support on CPUs still limits a large-scale adoption.…”