The sense of smell is one of the earliest senses formed in the human body, and it is a very complex sensory response. The nose is like a highly sensitive detector. Through millions of olfactory nerves, we are able to perceive and distinguish various odor molecules with different structural properties, so that we can make quick judgments in complex environments.
Image source: Pexel
With the continuous development of science and technology, artificial intelligence (AI) olfactory recognition technology that imitates human olfactory perception has developed rapidly. The technology combines advanced algorithms of machine learning and artificial intelligence to identify various substances by detecting and analyzing odor molecules. The application fields of AI olfactory technology range from environmental monitoring to medical diagnosis, from food safety to criminal investigation, and its potential is unlimited.
When AI in many fine fields has surpassed human capabilities, we can’t help but ask: In the “arena” of smell perception,AI or the human nose, which one has better smell perception?? Before uncovering the answer to this question, let’s first understand how humans and AI perceive smells.
How does the human brain perceive smell?
The process by which the brain perceives smell is like an "encounter."
First, the odor molecules arrive quietly and enter the nasal cavity. There is a special area above the nasal cavity called the olfactory epithelium.
There are a large number of olfactory receptor cells here that can specifically recognize odor molecules. These odor molecules travel around the nasal cavity looking for their partners, our olfactory receptors.
Humans have approximately 400 functional olfactory receptors. Once these receptors come into contact with odor molecules, they will immediately cause changes in electrical signals and perform an "electric shock dance" to transmit signals to the brain.
This signal travels through the olfactory nerve to a specific area of the brain called the olfactory bulb.
Red is the olfactory bulb, picture source: Reference [1]
In the olfactory bulb, these signals are further processed and analyzed. The information is then sent to brain areas associated with memory and emotion, such as the hippocampus and amygdala. The brain converts these signals into smell sensations that we can recognize and understand, allowing us to experience the taste, texture and other characteristics of the smell.
Finally, the processing of olfactory nerve signals forms semantic representations describing various smells, such as coffee, rose, durian, etc. This process is so magical and exquisite, filling our lives with color and fun.
How does AI "smell" smells?
We now have a general understanding of the principles and processes of how the human brain perceives smells, so how does AI smell various smells?
AI “smells” scents like a “guessing game” based on molecular structure.
Odor comes from molecules with specific structures, which are like "messengers" carrying odor signals. therefore,To predict the smell of a substance, the key is to identify the composition and structure of the molecules.
In this process, AI relies on a large and carefully organized database. This database can be thought of as an advanced "odor-molecule translation dictionary" with an exhaustive list of connections between known molecular structures and their corresponding odors. Each molecule's association with odor is meticulously recorded and archived.
A neural algorithm used to simulate biological smell is reported in a paper in the journal Nature Machine Intelligence
When faced with the task of predicting the odor of a new molecule, AI will quickly search this professional "dictionary" to find known molecules with similar structures to the new molecule, and infer possible odor attributes from them. Not only is this process fast, it's also extremely precise.
In addition to basic structural matching, AI will also comprehensively consider other chemical properties, such as the electronegativity and stereoconfiguration of molecules, to more comprehensively predict the odor characteristics of new molecules.
The overall process is like an AI gathering and analyzing various clues to deduce the likely smell of a new molecule.
In August 2023, a graph neural network (GNN) model for AI odor analysis was published in Science magazine.
The process of AI identifying odor, picture source: Reference [4]
After the molecular structure is input into the model, the GNN will optimize the weight of different chemical structures in specific odors, and finally judge the odor of the molecules through the prediction layer and output the corresponding odor descriptors.
The researchers conducted scent tests on the GNN model and on a group of humans. The results show,AI outperformed human experts on 53% of chemical molecules and 55% of odor descriptions.
Humans vs. AI: Who is the smell expert?
We can imagine a team of professionals who are “smell experts”.
Unlike AI, which relies on large amounts of data and algorithms, these experts rely primarily on their sense of smell and years of accumulated experience to parse and describe smells. They have the ability to recognize the nuances of a variety of complex odors and describe them with precise language.
For example, they can clearly distinguish between different types of smells such as floral, fruity, grassy, leathery, etc., and explain them in depth.
In addition, these odor experts are able to analyze and interpret odors in conjunction with their origin and environmental factors. For example, they can identify the smells produced by the cooking process, the smells of plants, the unique smells of animals, etc., and systematically analyze the causes and effects of these smells based on their characteristics and changes.
Unlike predictions from data-driven AI,The descriptions and judgments of these scent experts may be affected by subjective factors.
Their conclusions may vary from person to person and may even be affected by a variety of factors. Therefore, in some cases, their description of the smell may differ from the AI's judgment.
Of course, this is just an imagination and does not mean that real scent experts are not professional. At the current stage, AI's olfactory ability has not yet reached the level of overwhelming humans, and humans have irreplaceable advantages in subjective experience and understanding of smells.
First of all, in the face of complex odor spectrum, AI needs to rely on massive data and advanced algorithms for learning and simulation in order to output more accurate judgments.
However, the human olfactory system can show higher flexibility, which is still difficult for current AI systems to achieve.
Correlation of predictions from different AI models to the human group average. Image source: Reference [4]
The human sense of smell is also affected by many other factors, such as mood, health, life experience, etc. These factors may influence our perception and judgment of smells.
These variables add a layer of complexity to the human sense of smell that AI lacks to fully understand and simulate the human olfactory system.
Conclusion
Although AI has shown impressive potential in olfactory technology and has made significant progress in some fields,But it has not yet fully surpassed humans. Each has different advantages and limitations. As technology continues to advance, we have reason to expect that AI will achieve more breakthroughs in smell.
However, this field still faces various challenges, such as accurate identification, stability and reproducibility of odor molecules, which require further research and improvement. also,Public acceptance and trust in this emerging technology are also key factors driving its successful application.
To sum up, AI olfactory technology has broad development prospects and unlimited possibilities, but its specific development trajectory and results still need time and practice to verify. We look forward to scientific researchers and engineers being able to solve these challenges and bring more convenience and safety to society.
References
[1] Edmund Chong, Monica Moroni, Christopher Wilson, et al. Manipulating synthetic synthetic optogenetic odors reveals the coding logic of olfactory perception Science 2020, 368, 6497.
[2]LuluGuo,JieCheng,ShuoLian,etal.Structuralbasisofamineodorantperceptionbyamammalolfactoryreceptor.Nature2023,618,193.
[3]JiaDuan,PeiyuXu,XiaodongLuan,etal.Hormone-andantibody-mediatedactivationofthethyrotropinreceptor.Nature2022,609,854.
[4]BrianK.Lee,EmilyJ.Mayhew,BenjaminSanchez-Lengeling,etal.Aprincipalodormapunifiesdiversetasksinolfactoryperception.Science2023,381,999.