SOM-CNN
No estimate
No hardware requirements for this model
The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.
On record
Full specification
Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.
Origin
Who built this model, where, and when it was published.
- Published
- 31 January 1997
- Authors
- S. Lawrence, C. Lee Giles, A. Tsoi, A. Back
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Vision
- Task
- Face recognition
- Approach
- Supervised
Size
How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.
- Parameters
- 32K
- Training data
- 129,000 tokens
- Epochs
- 100
SOM: 125 L1: 20*3*3*3=540 L2: 25*3*3*20*0.3=1350 L3: 25*5*6*40=30000 Total: 125+540+1350+30000=32,015 (see Table 1)
"a total of 200 training images and 200 test images." but note that images are sub-sampled to smaller local windows for training.
Training compute
The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.
- Training compute
- 3.1 × 10¹⁰ FLOP
- How it was established
- Operation counting
Forward pass flop based on architecture (see Table 1): First CNN layer: 23*28 -> 3*3*20 2*21*26*3*3*20=196560 Second CNN layer: 20*11*13 -> 3*3*25*0.3 2*9*11*3*3*20*25*0.3=267300 FC layer: 25*5*6 -> 40 2*25*5*6*40=60000 Total forward FLOP: 196560+267300+60000=523860 Total training compute: 523860*3*20000=31431600000 Training steps: "The network was trained with backpropagation [13] for a total of 20 000 updates."
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Why it is tracked
- Historical significance,Highly cited
- Record confidence
- Confident
Sources
Where this record came from and when it was last checked.
- Reference
- Face recognition: a convolutional neural-network approach
- Last updated
- 28 November 2025
What the numbers mean
About this model
SOM-CNN was published by its authors, in January 1997.
It works in Vision, and is recorded as doing face recognition.
Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.
How it was trained
The training run consumed about 3.1 × 10¹⁰ FLOP. That figure describes the cost of creating it and has no bearing on how quickly it generates text.
Around 129,000 tokens went into training it.
It is tracked in the underlying dataset for one reason in particular: historical significance,Highly cited.
Answers
SOM-CNN — common questions
How many parameters does SOM-CNN have?
SOM-CNN has 32K parameters. SOM: 125 L1: 20*3*3*3=540 L2: 25*3*3*20*0.3=1350 L3: 25*5*6*40=30000 Total: 125+540+1350+30000=32,015 (see Table 1). That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.
When was SOM-CNN released?
SOM-CNN was published in January 1997. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
What is SOM-CNN used for?
SOM-CNN works in Vision, and is recorded as handling face recognition. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.
How much compute was used to train SOM-CNN?
Around 3.1 × 10¹⁰ FLOP. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.
What GPU do I need to run SOM-CNN?
None. SOM-CNN is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.
Is SOM-CNN open source?
The licensing for SOM-CNN was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.
The other direction
Looking at it from the other side?
This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.