icevision
Agnostic Computer Vision Framework
Decision gist · record as of 2026-08-14
No. The package is abandoned (no updates since February 2022) and depends on a large, rapidly evolving stack of computer vision libraries that may have become incompatible. For new projects, use actively maintained alternatives like Ultralytics YOLOv5/v8, Detectron2, or MMDetection. Install only if you are maintaining legacy code already using IceVision and cannot migrate.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Linux/MacOS only; requires PyTorch and torchvision installed; abandoned project with no active maintenance since 2022-02-10.
- Low install friction, but the package is abandoned as of 1646 days since its last release.
- Depends on 15 runtime packages including torch, torchvision, and several specialized computer vision libraries, which may require system-level dependencies or significant disk space.
License · maintenance · safety
Apache-2.0 (permissive) — Licensed under Apache-2.0 (permissive), allowing commercial use, modification, and distribution with minimal restrictions—suitable for most projects that can accept the maintenance risk.
last release 2022-02-10 (1646 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 473,813 downloads/mo, #6,458 on PyPI
Alternatives
Verify before relying
pip install icevision[all]
from icevision.all import *
# Load a pre-trained model and prepare data for training- Whether the package's dependencies (torch, torchvision, yolov5-icevision, effdet, sahi, resnest) remain compatible with current Python and PyTorch versions.
- Whether the model zoo and pre-trained weights referenced in the description are still accessible and functional.
- Whether the exploratory data analysis dashboard and auto-fix features work as documented.
What it is and what it does
IceVision is an object detection framework that abstracts away model and training backend differences, letting you work with hundreds of pre-trained models from torchvision, MMLabs, and other sources through a single API. It handles the full pipeline—data curation, augmentation, model selection, and training orchestration—using libraries like PyTorch Lightning and Fastai as pluggable backends.
The package includes data cleaning tools, an exploratory dashboard, pluggable transforms for augmentation, and support for multi-task learning combining detection, segmentation, and classification. However, the project has been abandoned since February 2022 and receives no active maintenance. It requires Linux or macOS and depends on a large stack of computer vision libraries (torch, torchvision, opencv-python, albumentations, and specialized detection models).
Use it for
- Train object detection models on custom datasets using pre-trained weights from a curated model zoo without writing boilerplate training code.
- Explore and clean image datasets using the built-in data curation and auto-fix tools before training.
- Combine object detection with image segmentation or classification in a single multi-task training pipeline.
- Experiment with different neural network architectures and training backends (PyTorch Lightning, Fastai) on the same dataset.
- Leverage hundreds of pre-trained models from torchvision and MMLabs for transfer learning on detection tasks.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
No.
The package is abandoned (no updates since February 2022) and depends on a large, rapidly evolving stack of computer vision libraries that may have become incompatible. For new projects, use actively maintained alternatives like Ultralytics YOLOv5/v8, Detectron2, or MMDetection. Install only if you are maintaining legacy code already using IceVision and cannot migrate.
Install
icevision on PyPI
Before you install
Low install friction, but the package is abandoned as of 1646 days since its last release. Depends on 15 runtime packages including torch, torchvision, and several specialized computer vision libraries, which may require system-level dependencies or significant disk space.
Linux/MacOS only; requires PyTorch and torchvision installed; abandoned project with no active maintenance since 2022-02-10.
License in practice
Licensed under Apache-2.0 (permissive), allowing commercial use, modification, and distribution with minimal restrictions—suitable for most projects that can accept the maintenance risk.
Quickstart
pip install icevision[all]
from icevision.all import *
# Load a pre-trained model and prepare data for trainingVerify before relying
- Whether the package's dependencies (torch, torchvision, yolov5-icevision, effdet, sahi, resnest) remain compatible with current Python and PyTorch versions.
- Whether the model zoo and pre-trained weights referenced in the description are still accessible and functional.
- Whether the exploratory data analysis dashboard and auto-fix features work as documented.
Package facts
| License | Apache-2.0 permissive |
| Python support | Supports the current Python release <4,>=3.7 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 15 packagespillowtorchtorchvisionfastcoretqdmopencv-pythonalbumentationsresnesteffdetsahiyolov5-icevisionipykerneldataclassesloguruimportlib-metadata |
| Maintenance | Abandoned 1,646 days since the last release |
| First released | |
| Downloads | 473,813 / month, #6,458 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 4 - BetaIntended Audience :: DevelopersProgramming Language :: Python :: 3.7Programming Language :: Python :: 3.8Topic :: Scientific/Engineering :: Artificial IntelligenceTopic :: Scientific/Engineering :: Image Recognition |
Evidence: icevision-0.12.0-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “multi-model detection pipeline”
- icevisionIceVision provides a unified framework for training and deploying…
- litserveLitServe lets you build custom inference servers in Python by…
- cosmos-xennaCosmos-xenna is a Python framework for building and running…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also pytorchcv · torchvision · lightning · ultralytics · sae-lens · autogluon.vision · mmcv · fvcore · fiftyone-db · modelscope