Skin Tone Estimation under Diverse Lighting Conditions.

Knowledge of a person's level of skin pigmentation, or so-called "skin tone", has proven to be an important building block in improving the performance and fairness of various applications that rely on computer vision. These include medical diagnosis of skin conditions, cosmetic and skincare support, and face recognition, especially for darker skin tones. However, the perception of skin tone, whether by the human eye or by an optoelectronic sensor, uses the reflection of light from the skin. The source of this light, or illumination, affects the skin tone that is perceived. This study aims to refine and assess a convolutional neural network-based skin tone estimation model that provides consistent accuracy across different skin tones under various lighting conditions. The 10-point Monk Skin Tone Scale was used to represent the skin tone spectrum. A dataset of 21,375 images was captured from volunteers across the pigmentation spectrum. Experimental results show that a regression model outperforms other models, with an estimated-to-target distance of 0.5. Using a threshold estimated-to-target skin tone distance of 2 for all lights results in average accuracy values of 85.45% and 97.16%. With the Monk Skin Tone Scale segmented into three groups, the lighter exhibits strong accuracy, the middle displays lower accuracy, and the dark falls between the two. The overall skin tone estimation achieves average error distances in the LAB space of 16.40±20.62.

Dataset of indoor air parameter measurements relating to indoor air quality and thermal comfort in South African primary school classrooms of various building infrastructure types.

In the resource-constrained South African education sector, infrastructure considered temporary or a backup in other countries is used as permanent classrooms, primarily but not exclusively in lower-income areas. Children's cognitive performance and comfort are directly impacted by indoor air quality. Temperature, relative humidity, particulate matter and CO2 levels, substantial determinants of air quality and thermal comfort, have not been investigated across different classroom building and infrastructure types. We measure these parameters with 11-min intervals in 24 classrooms at schools in Stellenbosch, South Africa. These classrooms consist of a range of different infrastructure types. Container classrooms with and without insulation, mobile (prefabricated) classrooms, and brick classrooms of different configurations are included. Measurements are concurrently sampled over ten months (249 days, still ongoing) across multiple seasons with relevant metadata, including ambient weather conditions, school days and times, and electricity availability in the (South) African context, which impacts air conditioning usage. This dataset provides valuable insights into true learning conditions in South African classrooms.

Dataset of temperature, humidity, and actuator states of an east-facing South African Greenhouse Tunnel.

A greenhouse tunnel in Stellenbosch, South Africa was used for testing a generic sensing system for monitoring and control of climatic conditions in the tunnel. Three temperature and humidity sensors were used to record data throughout the day in 5 min intervals. Bambara Nuts, a climate change-resilient and nutritious crop, were grown in a separate study in the tunnel using an aeroponics system. These were chosen as it is regarded as the norm in autonomous greenhouse temperature control in the region. During data collection, the sensors were placed at the front, middle, and back of the tunnel. At the front, there was an industrial extraction fan, and at the back, there was an evaporative cooling wet wall. The fan and wet wall were controlled using the middle sensor data that was averaged every minute to determine if the fan and wet wall should be on or off. The hysteresis band used as a threshold was to turn the fan on when the middle temperature reached 30 °C and to turn it off it was 22 °C. This data collection method extended from 31 December 2022 to 13 June 2023, collecting 162 days of temperature and humidity data for that period.

Using infrared to improve face recognition of individuals with highly pigmented skin.

Face recognition is widely used for security and access control. Its performance is limited when working with highly pigmented skin tones due to training bias caused by the under-representation of darker-skinned individuals in existing datasets and the fact that darker skin absorbs more light and therefore reflects less discernible detail in the visible spectrum. To improve performance, this work incorporated the infrared (IR) spectrum, which is perceived by electronic sensors. We augmented existing datasets with images of highly pigmented individuals captured using the visible, IR, and full spectra and fine-tuned existing face recognition systems to compare the performance of these three. We found a marked improvement in accuracy and AUC values of the receiver operating characteristic (ROC) curves when including the IR spectrum, increasing performance from 97.5% to 99.0% for highly pigmented faces. Different facial orientations and narrow cropping also improved performance, and the nose region was the most important feature for recognition.

First report of aeroponically grown Bambara groundnut, an African indigenous hypogeal legume: Implications for climate adaptation.

Global agricultural production is currently limited by negative climate-related hazards such as drought, uneven rainfall and rising temperatures. Many efforts have been put in place by government and non-government agencies to mitigate the challenges of climate change in the sector. However, the approaches do not seem feasible due to the growing demand for food. With these challenges, climate-smart agricultural technologies such as aeroponics and underutilised crops have been projected as the future of agriculture in developing African countries to reduce the risk of food insecurity. In this paper, we present the cultivation of an underutilised indigenous African legume crop, Bambara groundnut, in an aeroponics system. Seventy Bambara groundnut landraces were cultivated in a low-cost climate-smart aeroponics system and in sawdust media. The results showed that Bambara groundnut landraces cultivated in aeroponics performed better than those cultivated in a traditional hydroponics (sawdust/drip irrigation) technique in terms of plant height and chlorophyll content, where the landraces cultivated in sawdust had a higher number of leaves than those cultivated in aeroponics. This study also demonstrated the feasibility of introducing a generic Internet of Things platform for climate-smart agriculture in developing countries. The proof-of-concept and the successful cultivation of a hypogeal crop in aeroponics can be useful for cost-effective adaptation and mitigation plans for climate change, particularly for food security in rural African agricultural sectors.

Canary in the coliform mine: Exploring the industrial application limits of a microbial respiration alarm system.

Fundamental ecological principles of ecosystem-level respiration are extensively applied in greenhouse gas and elemental cycle studies. A laboratory system termed CEMS (Carbon Dioxide Evolution Measurement System), developed to explore microbial biofilm growth and metabolic responses, was evaluated as an early-warning system for microbial disturbances in industrial settings: in (a) potable water system contamination, and (b) bioreactor inhibition. Respiration was detected as CO2 production, rather than O2 consumption, including aerobic and anaerobic metabolism. Design, thresholds, and benefits of the remote CO2 monitoring technology were described. Headspace CO2 correlated with contamination levels, as well as chemical (R2 > 0.83-0.96) and microbiological water quality indicators (R2 > 0.78-0.88). Detection thresholds were limiting factors in monitoring drinking water to national and international standards (0 CFU/100 mL fecal coliforms) in both open- (>1500 CFU/mL) and closed-loop CO2 measuring regimes (>100 CFU/100 mL). However, closed-loop detection thresholds allow for the detection of significant contamination events, and monitoring less stringent systems such as irrigation water (<100 CFU/mL). Whole-system respiration was effectively harnessed as an early-warning system in bioreactor performance monitoring. Models were used to deconvolute biological CO2 fluctuations from chemical CO2 dynamics, to optimize this real-time, sustainable, low-waste technology, facilitating timeous responses to biological disturbances in bioreactors.