Hierarchical Scene Labeling (Stanford Background)

Closed weights New York University (NYU) 51.6M parameters August 2013

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.

Organisation
New York University (NYU)
Organisation type
Academia
Country
United States of America
Published
1 August 2013
Authors
Clement Farabet; Camille Couprie; Laurent Najman; Yann LeCun

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Vision
Task
Semantic segmentation
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
51.6M

Figure 1 details architecture, weights are shared between scales “The network is then applied to each 3-dimension input map Xs. This input is transformed into a 16-dimension feature map, using a bank of 16 filters, 10 connected to the Y channel, the 6 others connected to the U and V channels. The second layer transforms this 16-dimension feature map into a 64-dimension feature map, each map being produced by a combination of 8 randomly selected feature maps from the previous layer. Finally the 6…

Training data
71,380,800 tokens

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
2.4 × 10¹⁷ FLOP

Forward FLOP Split for the three sizes F1: L1: 2*1*7*7*10*320*240+2*2*7*7*6*320*240=165580800 L2: 2*8*64*7*7*160*120=963379200 FC: 2*32*80*60*256=78643200 F2: L1: 2*1*7*7*10*160*120+2*2*7*7*6*160*120=41395200 L2: 2*8*64*7*7*80*60=240844800 FC: 2*32*40*30*256=19660800 F3: L1: 2*1*7*7*10*80*60+2*2*7*7*6*80*60=10348800 L2: 2*8*64*7*7*40*30=60211200 FC: 2*32*20*15*256=4915200 Total: 165580800+963379200+78643200+41395200+240844800+19660800+10348800+60211200+4915200=1584979200 Training compute: 15849…

How it was established
Operation counting

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Chips used
1

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
Unreleased

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
Highly cited,Historical significance
Record confidence
Confident

Sources

Where this record came from and when it was last checked.

Reference
Learning Hierarchical Features for Scene Labeling
Last updated
11 February 2026

What the numbers mean

About this model

Hierarchical Scene Labeling (Stanford Background) was published by New York University (NYU), in United States of America, in August 2013. It comes out of academia.

It works in Vision, and is recorded as doing semantic segmentation.

This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.

How it was trained

Producing it required around 2.4 × 10¹⁷ FLOP of arithmetic, which is a statement about the training budget rather than about inference.

Around 71,380,800 tokens went into training it.

It is tracked in the underlying dataset for one reason in particular: highly cited,Historical significance.

Answers

Hierarchical Scene Labeling (Stanford Background) — common questions

01

Is Hierarchical Scene Labeling (Stanford Background) open source?

No. Hierarchical Scene Labeling (Stanford Background) has not had its weights published, so it exists only as a service controlled by its owner.

02

How many parameters does Hierarchical Scene Labeling (Stanford Background) have?

Hierarchical Scene Labeling (Stanford Background) has 51.6M parameters. Figure 1 details architecture, weights are shared between scales “The network is then applied to each 3-dimension input map Xs. This input is transformed into a 16-dimension feature map, using a bank of 16 filters, 10 connected to the Y channel, the 6 others connected to the U and V channels. The second layer transforms this 16-dimension feature map into a 64-dimension feature map, each map being produced by a combination of 8 randomly selected feature maps from the previous layer. Finally the 64-dimension feature map is transformed into a 256-dimension feature map, each map being produced by a combination of 32 randomly selected feature maps from the previous layer.” L1: 1*7*7*10+2*7*7*6=1078 L2: 8*64*7*7=25088 FC head is split for the three networks (assumption) FC: 32*80*60*256+32*40*30*256+32*20*15*256=51609600 This is a very different estimate to the 0.5M trainable parameters mentioned in the paper(!) "The convolutional network has roughly 0.5 million trainable parameters". 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.

03

Who created Hierarchical Scene Labeling (Stanford Background)?

Hierarchical Scene Labeling (Stanford Background) was published by New York University (NYU), based in United States of America, categorised as academia.

04

When was Hierarchical Scene Labeling (Stanford Background) released?

Hierarchical Scene Labeling (Stanford Background) was published in August 2013. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

05

What is Hierarchical Scene Labeling (Stanford Background) used for?

Hierarchical Scene Labeling (Stanford Background) works in Vision, and is recorded as handling semantic segmentation. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

06

How much compute was used to train Hierarchical Scene Labeling (Stanford Background)?

Around 2.4 × 10¹⁷ 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.

07

What GPU do I need to run Hierarchical Scene Labeling (Stanford Background)?

None. Hierarchical Scene Labeling (Stanford Background) 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.

Source

Original publication

Record last updated 11 February 2026

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.