Feasibility of ceiling-based luminance distribution measurements

Kruisselbrink, T.W.; Dangol, Rajendra; Loenen, E.J. van

doi:10.1016/j.buildenv.2020.106699

Cited by 16 publications

(21 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The learned self-supervised model is fine-tuned on a small labeled dataset to minimize the requirement of costly semantic annotations. We demonstrate that LumNet achieves a better performance than prior methods based on domain-knowledge [27,28] when trained on labeled data consisting of ceiling-based images. This also holds when LumNet representations learned from a labeled data are transferred for fine-tuning on a smaller dataset consisting of different VVF positions.…”

Section: Introductionmentioning

confidence: 95%

“…This corresponds to 135°, specifically 60°upwards and 70°downwards [52]. Systems which use these classical strategies have the main drawback that the cameras can be in the way, interfering with daily user activities and might also omit the user's direct task area [28]. The most suitable camera position has been shown to be on the ceiling, where the cameras are not obtrusive and can capture the direct task areas of all users.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

LumNet

Songwa

Saeed

Bhardwaj

et al. 2021

Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.

View full text Add to dashboard Cite

High-quality lighting positively influences visual performance in humans. The experienced visual performance can be measured using desktop luminance and hence several lighting control systems have been developed for its quantification. However, the measurement devices that are used to monitor the desktop luminance in existing lighting control systems are obtrusive to the users. As an alternative, ceiling-based luminance projection sensors are being used recently as these are unobtrusive and can capture the direct task area of a user. The positioning of these devices on the ceiling requires to estimate the desktop luminance in the user's vertical visual field, solely using ceiling-based measurements, to better predict the experienced visual performance of the user. For this purpose, we present LUMNET, an approach for estimating desktop luminance with deep models through utilizing supervised and self-supervised learning. Our model learns visual representations from ceiling-based images, which are collected in indoor spaces within the physical vicinity of the user to predict average desktop luminance as experienced in a real-life setting. We also propose a self-supervised contrastive method for pre-training LUMNET with unlabeled data and we demonstrate that the learned features are transferable onto a small labeled dataset which minimizes the requirement of costly data annotations. Likewise, we perform experiments on domain-specific datasets and show that our approach significantly improves over the baseline results from prior methods in estimating luminance, particularly in the low-data regime. LUMNET is an important step towards learning-based technique for luminance estimation and can be used for adaptive lighting control directly on-device thanks to its minimal computational footprint with an added benefit of preserving user's privacy.

show abstract

Section: Introductionmentioning

confidence: 95%

Section: Introductionmentioning

confidence: 99%

LumNet

Songwa

Saeed

Bhardwaj

et al. 2021

Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.

View full text Add to dashboard Cite

show abstract

“…We acknowledge that a light sensor will have a measurement error because it is located at a different position than the eyes, e.g. [32]. Moreover, illuminance (in lux) is not the best way to measure light that influences the circadian rhythm, as suggested in e.g.…”

Section: Input Measurement Noisementioning

confidence: 99%

“…The 'max'-operation is required to prevent negative values forÎ. Realistic values for this gain and offset have been determined for a ceiling-or wall mounted light sensor in [32], which for a calibrated device reports a bestcase non-image forming gain error of ±32% and an offset of ∼ 0.2%. To simulate such an error we will consider this offset negligible and will draw a gain for each particle, according to…”

Section: Input Measurement Noisementioning

confidence: 99%

“…As light input, we used the realistic light profile, described near (35). The figures show that the inaccuracies of a ceiling-mounted light sensor only have a moderate impact on the circadian state, particularly when the sensor is calibrated to report the specific retinal light exposure [32]. The impact of the gain error depends on the light level, in concordance with the logarithmic light function in (3): The effect of the error will be relatively smaller for high light levels (say 10 000 lx) than for low levels (say 500 lx).…”

Section: Input Measurement Noisementioning

confidence: 99%

See 1 more Smart Citation

Parameter Estimation in a Model of the Human Circadian Pacemaker Using a Particle Filter

Bonarius¹,

Papatsimpa

Linnartz³

2021

IEEE Trans. Biomed. Eng.

View full text Add to dashboard Cite

In the near future, real-time estimation of people's unique, precise circadian clock state has the potential to improve the efficacy of medical treatments and improve human performance on a broad scale. Human-centric lighting can bring circadian-rhythm support using biodynamic lighting solutions that sync light with the time of day. We investigate a method to improve the tracking of individual's circadian processes. Methods: In literature, the human circadian physiology has been mathematically modeled using ordinary differential equations, the state of which can be tracked via the signal processing concept of a Particle Filter. We show that substantial improvements can be made if the estimator not only tracks state variables, such as the phase and amplitude of the circadian pacemaker, but also fits specific model parameters to the individual. In particular, we optimize model parameter τx, which reflects the intrinsic period of the circadian pacemaker (τ). We show that both state and model parameters can be estimated based on minimally-invasive light exposure measurements and sleep-wake state observations. We also quantify the effect of inaccuracies in sensing. Results: We demonstrate improved performance by estimating τx for every individual, both with artificially created and human subject data. Prediction accuracy improves with every newly available observation. The estimated τx-s correlate well with the subjects' chronotypes, in a similar way as τ correlates. Conclusion: Our results show that individualizing the estimation of model parameters can improve circadian state estimation accuracy. Significance: These findings underscore the potential improvements in personalized models over "one-size fits all" approaches. Index Terms-circadian rhythm, particle filter, parameter estimation I. INTRODUCTION T HE timing in several physiological processes in humans, including the sleep/wake cycle, hormone secretion, and subjective alertness and performance is controlled by a biological clock, called the circadian pacemaker [1]. On its own, this pacemaker oscillates with a near-24h period, hence the name coming from the latin words "circa" and "diem" meaning "approximately" and "a day". But the pacemaker Manuscript received month day, year; revised month day, year; accepted month day, year. Date of publication month day, year; date of current version month day, year.

show abstract

Semi-automated luminance map re-projection via high dynamic range imaging and indoor space 3-D reconstruction

Kim

Tzempelikos

2021

Automation in Construction

View full text Add to dashboard Cite

Feasibility of ceiling-based luminance distribution measurements

Cited by 16 publications

References 16 publications

LumNet

LumNet

Parameter Estimation in a Model of the Human Circadian Pacemaker Using a Particle Filter

Semi-automated luminance map re-projection via high dynamic range imaging and indoor space 3-D reconstruction

Contact Info

Product

Resources

About