Fit for Purpose? Predicting Perception Performance Based on Past Experience

Gurău, Corina; Tong, Chi Hay; Posner, Ingmar

doi:10.1007/978-3-319-50115-4_40

Cited by 12 publications

(15 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Accounting for Uncertainty: An AIA that can predict its performance on di erent tasks can provide assurances about competence, predictability, and the situational normality of a given task. Several researchers have worked to improve this ability in visual classi cation [25,51,66,151]. For example, to ensure that visual classi ers don't fail silently in novel scenarios, Zhang et al [151] learned models of errors on training images to predict errors on test images.…”

Section: Commonmentioning

confidence: 99%

“Dave...I can assure you ...that it’s going to be all right ...” A Definition, Case for, and Survey of Algorithmic Assurances in Human-Autonomy Trust Relationships

Israelsen

Ahmed

2019

ACM Comput. Surv.

View full text Add to dashboard Cite

People who design, use, and are a ected by autonomous arti cially intelligent agents want to be able to trust such agents -that is, to know that these agents will perform correctly, to understand the reasoning behind their actions, and to know how to use them appropriately. Many techniques have been devised to assess and in uence human trust in arti cially intelligent agents. However, these approaches are typically ad hoc, and have not been formally related to each other or to formal trust models. is paper presents a survey of algorithmic assurances, i.e. programmed components of agent operation that are expressly designed to calibrate user trust in arti cially intelligent agents. Algorithmic assurances are rst formally de ned and classi ed from the perspective of formally modeled human-arti cially intelligent agent trust relationships. Building on these de nitions, a synthesis of research across communities such as machine learning, human-computer interaction, robotics, e-commerce, and others reveals that assurance algorithms naturally fall along a spectrum in terms of their impact on an agent's core functionality, with seven notable classes ranging from integral assurances (which impact an agent's core functionality) to supplemental assurances (which have no direct e ect on agent performance).Common approaches within each of these classes are identi ed and discussed; bene ts and drawbacks of di erent approaches are also investigated.

show abstract

Section: Commonmentioning

confidence: 99%

“Dave...I can assure you ...that it’s going to be all right ...” A Definition, Case for, and Survey of Algorithmic Assurances in Human-Autonomy Trust Relationships

Israelsen

Ahmed

2019

ACM Comput. Surv.

View full text Add to dashboard Cite

show abstract

“…Our second approach also considers similarity of appearance between current and previous traversals (at the expense of this ability to predict this far in advance), which ensures that prior observations that differ greatly from the current frame, owing to, for example, vastly different lighting or weather conditions, are removed from consideration in the prediction model. This paper augments and extends work by Gurau et al (2016), with a more comprehensive exposition, additional experimental analysis with different detection models, and extended discussion. The key contributions of this work are:…”

Section: Introductionmentioning

confidence: 57%

“…This paper augments and extends work by Gurau et al (2016), with a more comprehensive exposition, additional experimental analysis with different detection models, and extended discussion. The key contributions of this work are:…”

Section: Introductionmentioning

confidence: 57%

Learn from experience: Probabilistic prediction of perception performance to avoid failure

Gurău

Rao

Tong³

et al. 2017

The International Journal of Robotics Research

Self Cite

View full text Add to dashboard Cite

Despite significant advances in machine learning and perception over the past few decades, perception algorithms can still be unreliable when deployed in challenging time-varying environments. When these systems are used for autonomous decision-making, such as in self-driving vehicles, the impact of their mistakes can be catastrophic. As such, it is important to characterize the performance of the system and predict when and where it may fail in order to take appropriate action. While similar in spirit to the idea of introspection, this work introduces a new paradigm for predicting the likely performance of a robot's perception system based on past experience in the same workspace. In particular, we propose two models that probabilistically predict perception performance from observations gathered over time. While both approaches are place-specific, the second approach additionally considers appearance similarity when incorporating past observations. We evaluate our method in a classical decision-making scenario in which the robot must choose when and where to drive autonomously in 60 km of driving data from an urban environment. Results demonstrate that both approaches lead to fewer false decisions (in terms of incorrectly offering or denying autonomy) for two different detector models, and show that leveraging visual appearance within a state-of-the-art navigation framework increases the accuracy of our performance predictions.

show abstract

“…Statistical outlier detection also falls into this category. Krug et al, 2016;Pinto and Gupta, 2016;Morrison et al, 2020) and are increasingly used, for example, to predict the performance of perception and vision-based navigation systems (e.g., Gurȃu et al, 2016;Daftry et al, 2016;Dequaire et al, 2016).…”

Section: Opportunities and Future Directionsmentioning

confidence: 99%

From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligence

Roy¹,

Posner²,

Barfoot³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Machine learning has long since become a keystone technology, accelerating science and applications in a broad range of domains. Consequently, the notion of applying learning methods to a particular problem set has become an established and valuable modus operandi to advance a particular field. In this article we argue that such an approach does not straightforwardly extended to robotics -or to embodied intelligence more generally: systems which engage in a purposeful exchange of energy and information with a physical environment. In particular, the purview of embodied intelligent agents extends significantly beyond the typical considerations of main-stream machine learning approaches, which typically (i) do not consider operation under conditions significantly different from those encountered during training; (ii) do not consider the often substantial, long-lasting and potentially safety-critical nature of interactions during learning and deployment; (iii) do not require ready adaptation to novel tasks while at the same time (iv) effectively and efficiently curating and extending their models of the world through targeted and deliberate actions. In reality, therefore, these limitations result in learning-based systems which suffer from many of the same operational shortcomings as more traditional, engineering-based approaches when deployed on a robot outside a well defined, and often narrow operating envelope. Contrary to viewing embodied intelligence as another application domain for machine learning, here we argue that it is in fact a key driver for the advancement of machine learning technology. In this article our goal is to highlight challenges and opportunities that are specific to embodied intelligence and to propose research directions which may significantly advance the state-of-the-art in robot learning.

show abstract

Fit for Purpose? Predicting Perception Performance Based on Past Experience

Cited by 12 publications

References 19 publications

“Dave...I can assure you ...that it’s going to be all right ...” A Definition, Case for, and Survey of Algorithmic Assurances in Human-Autonomy Trust Relationships

“Dave...I can assure you ...that it’s going to be all right ...” A Definition, Case for, and Survey of Algorithmic Assurances in Human-Autonomy Trust Relationships

Learn from experience: Probabilistic prediction of perception performance to avoid failure

From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligence

Contact Info

Product

Resources

About