ConvNets have been successfully deployed in many commercial applications from OCR to video surveillance, but require large amounts of labeled training samples.
Mixed Partly true and partly false in roughly equal measure.
Factual: Statements about the present or the past, checked against the record: did they hold up?
The second judge would not resolve this claim; the verdict shown is the first judge's.
What would count as true or false
Judge as of 2010. The claim has two parts: commercial deployment of ConvNets across applications from OCR to video surveillance, and their need for large labelled training sets. True: evidence shows ConvNets were deployed commercially by 2010 in several applications, including OCR or check reading (for example, the NCR check-reading systems) and surveillance-type uses such as face or person detection, and that supervised ConvNets were widely understood to need large labelled datasets. Mostly true: both parts hold, but the surveillance example or the word 'many' is slightly overstated. Mixed: only one of the two parts is supported. Mostly false: commercial deployment was limited to one or two niche uses, so 'many' is a significant overstatement. False: there was no commercial ConvNet deployment by 2010.
Set before the research began.
Evidence (3)
-
E1 What are convolutional neural networks (CNN)? - TechTalks TechTalks (BDTechTalks) · 6 January 2020
“CNNs found a niche market in banking and postal services and banking, where they read zip codes on envelopes and digits on checks.”
-
E2 What are convolutional neural networks (CNN)? - TechTalks TechTalks (BDTechTalks) · 6 January 2020
“During training, the developers provide the neural network with a large dataset of images annotated with their corresponding classes (cat, dog, horse, etc.).”
-
E3 What are convolutional neural networks (CNN)? - TechTalks TechTalks (BDTechTalks) · 6 January 2020
“CNNs needed a lot of data and compute resources to work efficiently for large images.”
Reasoning
E1 is independent and says ConvNets found a niche commercial market reading zip codes and check digits for postal services and banks. That supports OCR deployment, but the quote gives no dates. None of the evidence mentions video surveillance or face or person detection, and E1 calls the market a "niche", so the evidence does not support "many" commercial applications. E2 and E3 support the second part of the claim: supervised ConvNets need large labelled datasets and a lot of data. Overall, the evidence supports only one part of the claim fully.
Cite this claim
Yann LeCun, “ConvNets have been successfully deployed in many commercial applications from OCR to video surveillance, but require large amounts of labeled training samples.” (said 2010). Verdict: mixed. Reliability API, claim c_b88b0c3f4641, method 0.1.0+c876030c, dataset release v1.1.2. https://epistacklabs.com/claims/c_b88b0c3f4641
This claim is in dataset release v1.1.2, with the same verdict.
Reply to this claim
More factual claims by Yann LeCun
-
Mostly true
“For starters, transformers are equivariant to permutations.”
Transformers are equivariant to permutations, as ConvNets are equivariant to translations.
-
Mostly true
“If it were the case, we would have AI systems that could teach themselves to drive a car in 20 hours of practice, like any 17 year-old.”
-
Mostly true
“The architecture is good, the architecture of the encoder is good, but the fact that you train the system to reconstruct images does not lead it to produce to long, good generic features of images.”
-
Mostly true
“LLMs can do none of those or they can only do them in a very primitive way and they don’t really understand the physical world.”
-
Mostly true
“No, there’s one thing that autoregressive LLMs or that LLMs in general, not just the autoregressive one, but including the bird style bidirectional ones, are exploiting and its self supervised…”
Verdicts are made automatically; how verdicts are made.