gitwtfhub

wtf is fpn_tensorflow?

allentdan/fpn_tensorflow — explained in plain English

Analysis updated 2026-07-18 · repo last pushed 2019-03-26

Jupyter NotebookAudience · researcherComplexity · 4/5DormantSetup · hard

TL;DR

A TensorFlow implementation of Feature Pyramid Networks for object detection, helping computers find and box both tiny and large objects in the same photo.

Mindmap

mindmap
  root((repo))
    What it does
      Detects objects
      Draws boxes
      Multi-scale analysis
    Tech stack
      TensorFlow
      Faster R-CNN
      ResNet backbone
    Use cases
      Prototype detectors
      Train on custom data
      Benchmark on COCO
    Audience
      Researchers
      CV engineers
    Setup
      Pre-trained models
      Multi-GPU training

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

Why would anyone build with this?

REASON 1

Download a pre-trained model and run it on your own images to detect and box objects immediately.

REASON 2

Train the network on a custom dataset to detect objects specific to your domain, like security footage or satellite photos.

REASON 3

Use multi-GPU training to speed up training a Feature Pyramid Network on a large dataset.

REASON 4

Benchmark this detector's accuracy against standard datasets like Pascal VOC and COCO as a baseline for your own research.

What's in the stack?

TensorFlowPythonJupyter NotebookResNet

How it stacks up

allentdan/fpn_tensorflowakshit-python-programmer/text-detection-using-neural-networkanil-matcha/matchacodes
Stars—0—
LanguageJupyter NotebookJupyter NotebookJupyter Notebook
Last pushed2019-03-26—2018-08-16
MaintenanceDormant—Dormant
Setup difficultyhardeasyeasy
Complexity4/52/52/5
Audienceresearchervibe coderdeveloper

Figures from each repo's GitHub metadata at analysis time.

How do you spin it up?

Difficulty · hard Time to first run · 1h+

Requires a GPU and TensorFlow setup, multi-GPU training needs multiple graphics cards for speed.

Yoink these prompts

Prompt 1
Help me set up this Feature Pyramid Network TensorFlow project and run inference on a sample image using the pre-trained model.
Prompt 2
Show me how to prepare a custom dataset in the format this repo expects so I can train the Feature Pyramid Network detector on my own images.
Prompt 3
Explain how the Feature Pyramid Network combines multi-scale features with Faster R-CNN in this codebase to detect both small and large objects.
Prompt 4
Write a script that evaluates this model's detection accuracy on the COCO dataset and prints precision and recall per object size.

Frequently asked questions

wtf is fpn_tensorflow?

A TensorFlow implementation of Feature Pyramid Networks for object detection, helping computers find and box both tiny and large objects in the same photo.

What language is fpn_tensorflow written in?

Mainly Jupyter Notebook. The stack also includes TensorFlow, Python, Jupyter Notebook.

Is fpn_tensorflow actively maintained?

Dormant — no commits in 2+ years (last push 2019-03-26).

How hard is fpn_tensorflow to set up?

Setup difficulty is rated hard, with roughly 1h+ to a first successful run.

Who is fpn_tensorflow for?

Mainly researcher.

View the repo → Decode another repo

This repo across BitVibe Labs

Don't trust strangers blindly. Verify against the repo.

Other repos in this lane