At What Time in the Day Are We Most Productive?


The other day, I found myself feeling exceptionally tired, not getting much work done even though it was 11:30 in the morning. I thought this strange because this seems like a time when I should have be most productive. I then experienced a burst of energy in the afternoon (no coffee or energy drinks). I wondered what the conne ction was and so I decided to explore some research. The results were not entirely what I expected, and even helped me re-adjust my day accordingly in order to be more efficient with my time.

Affect Intensity Estimation Using Multiple Modalities

AAAI Conferences

One of the challenges in affect recognition is accurate estimation of the emotion intensity level. This research proposes development of an affect intensity estimation model based on a weighted sum of classification confidence levels, displacement of feature points and speed of feature point motion. The parameters of the model were calculated from data captured using multiple modalities such as face, body posture, hand movement and speech. A preliminary study was conducted to compare the accuracy of the model with the annotated intensity levels. An emotion intensity scale ranging from 0 to 1 along the arousal dimension in the emotion space was used. Results indicated speech and hand modality significantly contributed in improving accuracy in emotion intensity estimation using the proposed model.


AITopics Original Links

In a very dense clump of trees there might be no direct local intensity maximum or just one for the whole group.One of the early experiments of an automatic tree counting system, was based on digital aerial colour infrared films and on the detection of local maxima and is reported in Blazquez (1989). The flight height was 610 m (focal length: 150 mm). The interpretation involved registration of an individual tree in relation to other trees in a citrus grove. The localisation of individual trees was thus based on local image intensity peaks of the tree crown, where several peaks per tree were sequentially numbered. In this early work, it was concluded, e.g., that large trees planted in hedges could not be counted very well, which was due to the previously explained problem, that the correspondence between the local intensity maxima and the individual trees on the ground was not complete.

Kumamoto shaken by magnitude 5 earthquake registering maximum of lower 6 on Japan's seismic intensity scale

The Japan Times

KUMAMOTO - A magnitude 5.0 earthquake, registering a maximum of lower 6 on Japan's shindo intensity scale, hit Kumamoto Prefecture at around 6:10 p.m. Thursday, the Meteorological Agency said. No tsunami warning was issued. Local municipalities as well as police and firefighters scrambled to gauge the impact of the quake, with the extent of possible damage not yet confirmed as of Thursday evening. According to the agency, the focus of the earthquake was about 10 kilometers deep and shaking registering seismic intensity of lower 6 was observed in the northern part of Kumamoto Prefecture. Shaking registering an intensity of 4 on the scale was seen in the southern part of Fukuoka Prefecture, while some regions in Saga, Nagasaki, Oita and Miyazaki prefectures observed tremors that registered an intensity of 3, the agency said.

Poisson intensity estimation with reproducing kernels Machine Learning

Despite the fundamental nature of the inhomogeneous Poisson process in the theory and application of stochastic processes, and its attractive generalizations (e.g. Cox process), few tractable nonparametric modeling approaches of intensity functions exist, especially when observed points lie in a high-dimensional space. In this paper we develop a new, computationally tractable Reproducing Kernel Hilbert Space (RKHS) formulation for the inhomogeneous Poisson process. We model the square root of the intensity as an RKHS function. Whereas RKHS models used in supervised learning rely on the so-called representer theorem, the form of the inhomogeneous Poisson process likelihood means that the representer theorem does not apply. However, we prove that the representer theorem does hold in an appropriately transformed RKHS, guaranteeing that the optimization of the penalized likelihood can be cast as a tractable finite-dimensional problem. The resulting approach is simple to implement, and readily scales to high dimensions and large-scale datasets.