Wow, this is literally exactly what I was looking for. Thank you. Hopefully, it works well.
From https://developer.clarifai.com/quick-start/
Seems simple to train
// add inputs with concepts
app.inputs.create([{
"url": "https://samples.clarifai.com/dog1.jpeg",
"concepts": [
{ "id": "cat", "value": false },
{ "id": "dog", "value": true }
]
}, {
"url": "https://samples.clarifai.com/dog2.jpeg",
"concepts": [
{ "id": "cat", "value": false },
{ "id": "dog", "value": true }
]
Then you predict what the another image is: // predict the contents of an image by passing in a url
app.models.predict(Clarifai.GENERAL_MODEL,
'https://samples.clarifai.com/metro-north.jpg').then(
function(response) {
console.log(response);
},
function(err) {
console.error(err);
}
);I'm glad I could help! I'd actually be really curious to hear how well it works for you.
I've had success with Clarifai's [0] custom CV model API in the past. You basically upload batches of labeled images to train a model, and then you can submit new images for classification.
Of course, I have no idea how effective it would be for your documents. Obviously it depends on how visually distinct the different types are.
[0] https://www.clarifai.com