SAF R-CNN

Closed weights Beijing Institute of Technology,Sun Yat-sen University,Panasonic R&D,National University of Singapore 138M parameters October 2015

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
Beijing Institute of Technology,Sun Yat-sen University,Panasonic R&D,National University of Singapore
Organisation type
Academia,Academia,Industry,Academia
Country
China, Singapore
Published
28 October 2015
Authors
Jianan Li, Xiaodan Liang, ShengMei Shen, Tingfa Xu, Jiashi Feng, Shuicheng Yan

What it does

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

Domain
Vision
Task
Object detection
Approach
Supervised
Base model
VGG16

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
138M

Taken from VGG16 base model, ignoring minor architectural changes

Training data
350,000 tokens

"There are totally 350,000 bounding boxes of about 2,300 unique pedestrians labeled in 250,000 frames" "We use dense sampling of the training data (every 4th frame"

Epochs
7

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
1.2 × 10¹⁹ FLOP

Base model: 12291000000000002000 Base model inference FLOP: 15300000000 Finetune for 7 epochs with 62500 training examples 15300000000*3*62500*7=20081250000000000=2e16 Total compute: 12291000000000002000+20081250000000000=12311081250000002000

How it was established
Operation counting
Fine-tuning compute
2 × 10¹⁶ FLOP

15300000000*3*62500*7=20081250000000000=2e16

The training run

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

Training hardware
NVIDIA GeForce GTX TITAN X
Chips used
1
Power draw
291 W

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.

Frontier model
Yes
Why it is tracked
SOTA improvement

Pedestrian Detection on Caltech benchmark "our method achieves state-of-the-art performance on Caltech, INRIA, and ETH, and obtains competitive results on KITTI. "

Record confidence
Likely

Sources

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

Reference
Scale-aware Fast R-CNN for Pedestrian Detection
Last updated
28 November 2025

What the numbers mean

What this model is

SAF R-CNN was published by Beijing Institute of Technology,Sun Yat-sen University,Panasonic R&D,National University of Singapore, in China, in October 2015. The organisation is categorised as academia,Academia,Industry,Academia.

It works in Vision, and is recorded as doing object detection.

Its starting point was VGG16 — most models at this scale are adapted from an existing base rather than built from nothing.

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

What went into building it

Producing it required around 1.2 × 10¹⁹ FLOP of arithmetic, on NVIDIA GeForce GTX TITAN X, which is a statement about the training budget rather than about inference.

The training set ran to roughly 350,000 tokens.

Its inclusion criterion is sOTA improvement.

Answers

SAF R-CNN — common questions

01

Is SAF R-CNN open source?

No. SAF R-CNN has not had its weights published, so it exists only as a service controlled by its owner.

02

How many parameters does SAF R-CNN have?

SAF R-CNN has 138M parameters. Taken from VGG16 base model, ignoring minor architectural changes. 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 SAF R-CNN?

SAF R-CNN was published by Beijing Institute of Technology,Sun Yat-sen University,Panasonic R&D,National University of Singapore, based in China, categorised as academia,Academia,Industry,Academia.

04

When was SAF R-CNN released?

SAF R-CNN was published in October 2015. 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 SAF R-CNN used for?

SAF R-CNN works in Vision, and is recorded as handling object detection. These are the areas it was designed around; they describe intent rather than a hard boundary.

06

How much compute was used to train SAF R-CNN?

Around 1.2 × 10¹⁹ FLOP, on NVIDIA GeForce GTX TITAN X. 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 SAF R-CNN?

None. SAF R-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.

Source

Original publication

Record last updated 28 November 2025

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.