Could AI help winegrowers better predict grape harvests? Photo / File
Could AI help winegrowers better predict grape harvests? Photo / File
Kiwi researchers want to see whether artificial intelligence can do something seasoned winemakers have been trying to do for thousands of years: predicting the harvest.
"Grape growers and wineries spend a lot of money trying to predict their grape yield each year," explained Lincoln Agritech's Jaco Fourie, an image-processing expert who is involved in the new joint study.
"This currently involves hiring a large number of workers to manually sample grape bunches."
Lincoln Agritech is working on creating a more convenient system that uses electronic sensors to accurately count grapes.
"The sensors will capture and analyse grape bunches within individual rows, and assess the number, sizes and distribution of grape bunches," Fourie said.
"We'll then feed this data into computer algorithms, which have been designed by the University of Canterbury, to predict grape yield at harvest time."
New data will be added to the system each year, leading to continuous improvements in the model's accuracy, with the system's predictive power improving over time as more data is gathered under different conditions.
Profitable wine production depended on early knowledge of the grape yield that is likely to be harvested each season, he said.
"Estimating the yield as soon as possible allows marketers to know how much wine will end up being produced."
The main focus of grape varieties for the study, jointly funded by the Government and NZ Winegrowing, was sauvignon blanc, with the team then moving to pinot noir.
Signs of psychopathy in kids?
New research has shown boys at risk of developing psychopathy don't feel the urge to join in laughing with others. Photo / 123RF
For most people, laughter is highly contagious.
It's nearly impossible to hear or see someone laughing and not feel the urge to join in.
But researchers have new evidence to show that boys at risk of developing psychopathy when they become adults don't have that same urge.
Individuals at risk of psychopathy show persistent disruptive behaviours alongside callous-unemotional traits.
When asked in the UK-led study, boys fitting that description reported that they didn't want to join in with laughter as much as their peers.
Images of their brains also showed reduced response to the sound of laughter.
Those differences were seen in brain areas that promote joining in with others and resonating with other people's emotions, not in auditory brain areas.
"We wanted to investigate how boys at risk of developing psychopathy process emotions that promote social affiliation, such as laughter," said study leader Professor Essi Viding, of University College London.
Viding and colleagues recruited 62 boys aged 11 to 16 with disruptive behaviours with or without callous-unemotional traits and 30 normally behaved, matched controls.
The groups were matched on ability, socio-economic background, ethnicity, and handedness.
Viding said the findings showed that kids who were vulnerable to developing psychopathy don't experience the world quite like the rest of us.
"Those social cues that automatically give us pleasure or alert us to someone's distress do not register in the same way for these children," she said.
"That does not mean that these children are destined to become antisocial or dangerous; rather, these findings shed new light on why they often make different choices from their peers.
"We are only now beginning to develop an understanding of how the processes underlying prosocial behaviour might differ in these children."
Pigeons aren't so bird-brained after all. Photo / File
Think you're good at multi-tasking?
Those pesky pigeons that coat our cars and streets in muck are probably better at it.
But far from being bird-brained, pigeons are capable of switching between two tasks as quickly as humans - and even more quickly in certain situations.
These are the findings of biopsychologists who had performed the same behavioural experiments to test birds and humans.
The authors hypothesise that the cause of the slight multi-tasking advantage in birds is their higher neuronal density.
"For a long time, scientists used to believe the mammalian cerebral cortex to be the anatomical cause of cognitive ability; it is made up of six cortical layers," explained Dr Sara Letzner, from Germany's Ruhr-University Bochum.
"That means the structure of the mammalian cortex cannot be decisive for complex cognitive functions such as multi-tasking."
The pallium of birds did not have any layers comparable to those in the human cortex - but its neurons were more densely packed than in the cerebral cortex in humans.
Pigeons, for example, had six times as many nerve cells as humans per cubic millimetre of brain.
Consequently, the average distance between two neurons in pigeons was 50 per cent shorter than in humans.
As the speed at which nerve cell signals were transmitted is the same in both birds and mammals, researchers had assumed that information was processed more quickly in avian brains than in mammalian brains.
They tested this hypothesis using a multi-tasking exercise that was performed by 15 humans and 12 pigeons.
In the experiment, both the human and the avian participants had to stop a task in progress and switch over to an alternative task as quickly as possible.
The researchers had assumed that pigeons must have an advantage over humans because of their greater nerve cell density.
Yet in one test, they were stunned to find the birds were an incredible 250 milliseconds faster than humans.
"Researchers in the field of cognitive neuroscience have been wondering for a long time how it was possible that some birds, such as crows or parrots, are smart enough to rival chimpanzees in terms of cognitive abilities, despite their small brains and their lack of a cortex," Letzner said.
The results of the current study offered a partial answer to the mystery: it was precisely because their small, nerve cell-packed brain that birds were able to think faster.
That study followed Kiwi and German findings last year that showed how pigeons could learn to distinguish real words from non-words by visually processing their letter combinations.
In that experiment, pigeons were trained to peck four-letter English words as they came up on a screen, or to instead peck a symbol when a four-letter non-word, such as "URSP" was displayed.
The researchers added words one by one with the four pigeons in the study eventually building vocabularies ranging from 26 to 58 words and over 8000 non-words.
They put their performance on a par with that previously reported in baboons for this type of complex task.
