skip to main content


Title: Characterizing SLAM Benchmarks and Methods for the Robust Perception Age
The diversity of SLAM benchmarks affords extensive testing of SLAM algorithms to understand their performance, individually or in relative terms. The ad-hoc creation of these benchmarks does not necessarily illuminate the particular weak points of a SLAM algorithm when performance is evaluated. In this paper, we propose to use a decision tree to identify challenging benchmark properties for state-of-the-art SLAM algorithms and important components within the SLAM pipeline regarding their ability to handle these challenges. Establishing what factors of a particular sequence lead to track failure or degradation relative to these characteristics is important if we are to arrive at a strong understanding for the core computational needs of a robust SLAM algorithm. Likewise, we argue that it is important to profile the computational performance of the individual SLAM components for use when benchmarking. In particular, we advocate the use of time-dilation during ROS bag playback, or what we refer to as slo-mo playback. Using slo-mo to benchmark SLAM instantiations can provide clues to how SLAM implementations should be improved at the computational component level. Three prevalent VO/SLAM algorithms and two low-latency algorithms of our own are tested on selected typical sequences, which are generated from benchmark characterization, to further demonstrate the benefits achieved from computationally efficient components.  more » « less
Award ID(s):
1816138
NSF-PAR ID:
10111563
Author(s) / Creator(s):
; ;
Date Published:
Journal Name:
ICRA Workshop on SLAM Benchmarking
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. We consider a variant of the vehicle routing problem (VRP) where each customer has a unit demand and the goal is to minimize the total cost of routing a fleet of capacitated vehicles from one or multiple depots to visit all customers. We propose two parallel algorithms to efficiently solve the column-generation-based linear-programming relaxation for this VRP. Specifically, we focus on algorithms for the “pricing problem,” which corresponds to the resource-constrained elementary shortest path problem. The first algorithm extends the pulse algorithm for which we derive a new bounding scheme on the maximum load of any route. The second algorithm is based on random coloring from parameterized complexity which can be also combined with other techniques in the literature for improving VRPs, including cutting planes and column enumeration. We conduct numerical studies using VRP benchmarks (with 50–957 nodes) and instances of a medical home care delivery problem using census data in Wayne County, Michigan. Using parallel computing, both pulse and random coloring can significantly improve column generation for solving the linear programming relaxations and we can obtain heuristic integer solutions with small optimality gaps. Combining random coloring with column enumeration, we can obtain improved integer solutions having less than 2% optimality gaps for most VRP benchmark instances and less than 1% optimality gaps for the medical home care delivery instances, both under a 30-minute computational time limit. The use of cutting planes (e.g., robust cuts) can further reduce optimality gaps on some hard instances, without much increase in the run time. Summary of Contribution: The vehicle routing problem (VRP) is a fundamental combinatorial problem, and its variants have been studied extensively in the literature of operations research and computer science. In this paper, we consider general-purpose algorithms for solving VRPs, including the column-generation approach for the linear programming relaxations of the integer programs of VRPs and the column-enumeration approach for seeking improved integer solutions. We revise the pulse algorithm and also propose a random-coloring algorithm that can be used for solving the elementary shortest path problem that formulates the pricing problem in the column-generation approach. We show that the parallel implementation of both algorithms can significantly improve the performance of column generation and the random coloring algorithm can improve the solution time and quality of the VRP integer solutions produced by the column-enumeration approach. We focus on algorithmic design for VRPs and conduct extensive computational tests to demonstrate the performance of various approaches. 
    more » « less
  2. Obeid, I. ; Selesnik, I. ; Picone, J. (Ed.)
    The Neuronix high-performance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from low-end consumer grade devices such as the Nvidia GTX 970 to higher-end devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely compute-intensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process. Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floating-point calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should produce identical results. (3) A job should produce comparable results if the data is presented in a different order. System optimization requires an ability to directly compare error rates for algorithms evaluated under comparable operating conditions. However, it is a difficult task to exactly reproduce the results for large, complex deep learning systems that often require more than a trillion calculations per experiment [5]. This is a fairly well-known issue and one we will explore in this poster. Researchers must be able to replicate results on a specific data set to establish the integrity of an implementation. They can then use that implementation as a baseline for comparison purposes. A lack of reproducibility makes it very difficult to debug algorithms and validate changes to the system. Equally important, since many results in deep learning research are dependent on the order in which the system is exposed to the data, the specific processors used, and even the order in which those processors are accessed, it becomes a challenging problem to compare two algorithms since each system must be individually optimized for a specific data set or processor. This is extremely time-consuming for algorithm research in which a single run often taxes a computing environment to its limits. Well-known techniques such as cross-validation [5,6] can be used to mitigate these effects, but this is also computationally expensive. These issues are further compounded by the fact that most deep learning algorithms are susceptible to the way computational noise propagates through the system. GPUs are particularly notorious for this because, in a clustered environment, it becomes more difficult to control which processors are used at various points in time. Another equally frustrating issue is that upgrades to the deep learning package, such as the transition from TensorFlow v1.9 to v1.13, can also result in large fluctuations in error rates when re-running the same experiment. Since TensorFlow is constantly updating functions to support GPU use, maintaining an historical archive of experimental results that can be used to calibrate algorithm research is quite a challenge. This makes it very difficult to optimize the system or select the best configurations. The overall impact of all of these issues described above is significant as error rates can fluctuate by as much as 25% due to these types of computational issues. Cross-validation is one technique used to mitigate this, but that is expensive since you need to do multiple runs over the data, which further taxes a computing infrastructure already running at max capacity. GPUs are preferred when training a large network since these systems train at least two orders of magnitude faster than CPUs [7]. Large-scale experiments are simply not feasible without using GPUs. However, there is a tradeoff to gain this performance. Since all our GPUs use the NVIDIA CUDA® Deep Neural Network library (cuDNN) [8], a GPU-accelerated library of primitives for deep neural networks, it adds an element of randomness into the experiment. When a GPU is used to train a network in TensorFlow, it automatically searches for a cuDNN implementation. NVIDIA’s cuDNN implementation provides algorithms that increase the performance and help the model train quicker, but they are non-deterministic algorithms [9,10]. Since our networks have many complex layers, there is no easy way to avoid this randomness. Instead of comparing each epoch, we compare the average performance of the experiment because it gives us a hint of how our model is performing per experiment, and if the changes we make are efficient. In this poster, we will discuss a variety of issues related to reproducibility and introduce ways we mitigate these effects. For example, TensorFlow uses a random number generator (RNG) which is not seeded by default. TensorFlow determines the initialization point and how certain functions execute using the RNG. The solution for this is seeding all the necessary components before training the model. This forces TensorFlow to use the same initialization point and sets how certain layers work (e.g., dropout layers). However, seeding all the RNGs will not guarantee a controlled experiment. Other variables can affect the outcome of the experiment such as training using GPUs, allowing multi-threading on CPUs, using certain layers, etc. To mitigate our problems with reproducibility, we first make sure that the data is processed in the same order during training. Therefore, we save the data from the last experiment and to make sure the newer experiment follows the same order. If we allow the data to be shuffled, it can affect the performance due to how the model was exposed to the data. We also specify the float data type to be 32-bit since Python defaults to 64-bit. We try to avoid using 64-bit precision because the numbers produced by a GPU can vary significantly depending on the GPU architecture [11-13]. Controlling precision somewhat reduces differences due to computational noise even though technically it increases the amount of computational noise. We are currently developing more advanced techniques for preserving the efficiency of our training process while also maintaining the ability to reproduce models. In our poster presentation we will demonstrate these issues using some novel visualization tools, present several examples of the extent to which these issues influence research results on electroencephalography (EEG) and digital pathology experiments and introduce new ways to manage such computational issues. 
    more » « less
  3. Abstract Benchmark datasets and benchmark problems have been a key aspect for the success of modern machine learning applications in many scientific domains. Consequently, an active discussion about benchmarks for applications of machine learning has also started in the atmospheric sciences. Such benchmarks allow for the comparison of machine learning tools and approaches in a quantitative way and enable a separation of concerns for domain and machine learning scientists. However, a clear definition of benchmark datasets for weather and climate applications is missing with the result that many domain scientists are confused. In this paper, we equip the domain of atmospheric sciences with a recipe for how to build proper benchmark datasets, a (nonexclusive) list of domain-specific challenges for machine learning is presented, and it is elaborated where and what benchmark datasets will be needed to tackle these challenges. We hope that the creation of benchmark datasets will help the machine learning efforts in atmospheric sciences to be more coherent, and, at the same time, target the efforts of machine learning scientists and experts of high-performance computing to the most imminent challenges in atmospheric sciences. We focus on benchmarks for atmospheric sciences (weather, climate, and air-quality applications). However, many aspects of this paper will also hold for other aspects of the Earth system sciences or are at least transferable. Significance Statement Machine learning is the study of computer algorithms that learn automatically from data. Atmospheric sciences have started to explore sophisticated machine learning techniques and the community is making rapid progress on the uptake of new methods for a large number of application areas. This paper provides a clear definition of so-called benchmark datasets for weather and climate applications that help to share data and machine learning solutions between research groups to reduce time spent in data processing, to generate synergies between groups, and to make tool developments more targeted and comparable. Furthermore, a list of benchmark datasets that will be needed to tackle important challenges for the use of machine learning in atmospheric sciences is provided. 
    more » « less
  4. SUMMARY

    Combining finite element methods for the incompressible Stokes equations with particle-in-cell methods is an important technique in computational geodynamics that has been widely applied in mantle convection, lithosphere dynamics and crustal-scale modelling. In these applications, particles are used to transport along properties of the medium such as the temperature, chemical compositions or other material properties; the particle methods are therefore used to reduce the advection equation to an ordinary differential equation for each particle, resulting in a problem that is simpler to solve than the original equation for which stabilization techniques are necessary to avoid oscillations.

    On the other hand, replacing field-based descriptions by quantities only defined at the locations of particles introduces numerical errors. These errors have previously been investigated, but a complete understanding from both the theoretical and practical sides was so far lacking. In addition, we are not aware of systematic guidance regarding the question of how many particles one needs to choose per mesh cell to achieve a certain accuracy.

    In this paper we modify two existing instantaneous benchmarks and present two new analytic benchmarks for time-dependent incompressible Stokes flow in order to compare the convergence rate and accuracy of various combinations of finite elements, particle advection and particle interpolation methods. Using these benchmarks, we find that in order to retain the optimal accuracy of the finite element formulation, one needs to use a sufficiently accurate particle interpolation algorithm. Additionally, we observe and explain that for our higher-order finite-element methods it is necessary to increase the number of particles per cell as the mesh resolution increases (i.e. as the grid cell size decreases) to avoid a reduction in convergence order.

    Our methods and results allow designing new particle-in-cell methods with specific convergence rates, and also provide guidance for the choice of common building blocks and parameters such as the number of particles per cell. In addition, our new time-dependent benchmark provides a simple test that can be used to compare different implementations, algorithms and for the assessment of new numerical methods for particle interpolation and advection. We provide a reference implementation of this benchmark in aspect (the ‘Advanced Solver for Problems in Earth’s ConvecTion’), an open source code for geodynamic modelling.

     
    more » « less
  5. The massive surge in the amount of observational field data demands richer and more meaningful collab-oration between data scientists and geoscientists. This document was written by members of the Working Group on Case Studies of the NSF-funded RCN on Intelli-gent Systems Research To Support Geosciences (IS-GEO, https:// is-geo.org/ ) to describe our vision to build and enhance such collaboration through the use of specially-designed benchmark datasets. Benchmark datasets serve as summary descriptions of problem areas, providing a simple interface between disciplines without requiring extensive background knowledge. Benchmark data intend to address a number of overarching goals. First, they are concrete, identifiable, and public, which results in a natural coordination of research efforts across multiple disciplines and institutions. Second, they provide multi-fold opportunities for objective comparison of various algorithms in terms of computational costs, accuracy, utility and other measurable standards, to address a particular question in geoscience. Third, as materials for education, the benchmark data cultivate future human capital and interest in geoscience problems and data science methods. Finally, a concerted effort to produce and publish benchmarks has the potential to spur the development of new data science methods, while provid-ing deeper insights into many fundamental problems in modern geosciences. That is, similarly to the critical role the genomic and molecular biology data archives serve in facilitating the field of bioinformatics, we expect that the proposed geosciences data repository will serve as “catalysts” for the new discicpline of geoinformatics. We describe specifications of a high quality geoscience bench-mark dataset and discuss some of our first benchmark efforts. We invite the Climate Informatics community to join us in creating additional benchmarks that aim to address important climate science problems. 
    more » « less