Neuromorphic Odor Translator Helps Robots Express Their Feelings

Neural network trained to properly name organic molecules

Aug 2021, phys.org

https://techxplore.com/news/2021-07-neural-network-properly-molecules.html

Do you ever have a problem naming that smell? It's not just you. Science has this problem too, but maybe not for long. 

Smells are volatile organic compounds that have evaporated and entered your nose. Although they almost always occur in combination with others and not in isolation, us humans want to reduce smells to their individual components, and then name them. After all, in order to think about something, you have to know it's name. (Is that true?) 

The problem is that organic molecules are big, with lots of chemicals joined together in lots of ways, so coming up with a naming convention for all these permutations is hard. IUPAC, the International Union of Pure and Applied Chemistry, sets the convention for naming molecules. And boy is it complicated.

Take sugar, a common molecule known to us by its simple name "sucrose;" in IUPAC, it's called (2R,3R,4S,5S,6R)-2-[(2S,3S,4S, 5R)-3,4-dihydroxy-2,5-bis(hydroxymethyl)oxolan-2-yl]oxy-6-(hydroxymethyl)oxane-3,4,5-triol.

Since we do live in the computer age, folks want to automate this naming process for when they discover new molecules. But as you can imagine by looking at the IUPAC name for sucrose, the algorithm at the core of that naming convention is really hard to write. So they decided to use a neural network instead.*

*I'm casually calling a neural network "neuromorphic," but in the past few years, real neuromorphic computers have forced a distinction here that I'm ignoring here for the sake of a more clickable title. 

This new artificially intelligent chemical translator is not a magical structure-to-name translator that can just look at a chemical and give it a name; that's still out of reach. It does, however, translate between IUPAC and another naming convention called SMILES.

Trained on PubChem's 100 million molecules, this translator ultimately shows how the utility of the new approach of using neural networks to help us write algorithms from the bottom up instead of the top down, which really is a revolution in computing. 

And if you think it would be cool to have robots that can smell, or to ensure that future humans maintain their sense of smell as they evolve into hyperdimensional algorithms, then making odors machine-readable is how you do that.

via Skolkovo Institute of Science and Technology, Lomonosov Moscow State University and start-up Syntelly: Lev Krasnov et al, Transformer-based artificial neural networks for the conversion between chemical notations, Scientific Reports (2021). DOI: 10.1038/s41598-021-94082-y

Identifying the Smell-Language Interface

Olfactory Artist Peter De Cupere

Re: Research investigating the physiological basis for odor naming via event-related potentials (ERPs) and fMRI.

The scientists fed subjects cues, either visual or olfactory, followed by words either matching or not matching the cue. A picture of a rose followed by the word “Rose”, the scent of a rose followed by the word “Rose”, or maybe the word “Lemon” instead. What parts of the brain light up when they recognize the match or the mismatch?

The results show that the “cue-modality”, whether it was visual or olfactory, affected different areas of the brain. Performance in recognizing either a match or a mismatch was slower when presented with an olfactory cue versus a visual cue. It was also discovered that the same areas responsible for recognition of an olfactory cue-word match lit up before the word appeared, suggesting less ‘flexibility’ in semantic identification of odors.

Furthermore, when the word is presented for match validation, the cue is reactivated, or re- experienced. But for olfaction, the entire olfactory perception system is not activated, only the parts which had initially coded the sensation semantically. This echoes the assertion that smells cannot be “imagined” in the same way as visual stimuli.

Let's not forget that smell originally functioned as an automatic system with no intervention of cortical processing. Activate - inhibit, that is the way of chemo-sensation. The buck does not cognize the scent of the doe, it reacts. Most of our models or analogies for thinking are visually based. The interface between olfaction and language is akin to the inner mental space in its entirety. The Olfactory-Language Interface, on the other hand, is more like a short cut through this mental space. There is no time for deliberation against the simulated perception, such a thing was impossible or unknown to our organic ancestors.

The chemically-sensitive organism (a redundancy in itself), whether plant or animal, is tied to its environment. The separation between the body and the environment is ultimately what we call this mind space, and it is the thing that makes us human.

(olfactory literature double whammy)

excerpt from The Foul and the Fragrant: Odor and the French Social Imagination, Alain Corbin, 1988; quoted in Sandalwood and Carrion: Smell in Indian Religion and Culture, James McHugh, 2012:

A few simple stereotypes demonstrate the paradoxical nature of the sense of smell. Olfaction as the sense of lust, desire, and impulsiveness is associated with sensuality. Smelling and sniffing are associated with animal behavior. If olfaction were his most important sense, man's linguistic incapacity to describe olfactory sensations would turn him into a creature tied to his environment. Because they are ephemeral, olfactory sensations can never provide a persistent stimulus of thought. Thus the development of the sense of smell seems to be inversely related to the development of intelligence.