{"id":20786338,"url":"https://github.com/praeclarumjj3/backbone-profile","last_synced_at":"2026-05-17T02:05:15.656Z","repository":{"id":69335389,"uuid":"323190633","full_name":"praeclarumjj3/BackBone-Profile","owner":"praeclarumjj3","description":"Inference Time Performance stats for various backbone networks.","archived":false,"fork":false,"pushed_at":"2020-12-28T15:24:50.000Z","size":182,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-10-29T02:40:59.005Z","etag":null,"topics":["inference-performance","pytorch"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/praeclarumjj3.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2020-12-21T00:10:13.000Z","updated_at":"2020-12-28T15:27:31.000Z","dependencies_parsed_at":"2023-04-10T09:01:51.891Z","dependency_job_id":null,"html_url":"https://github.com/praeclarumjj3/BackBone-Profile","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/praeclarumjj3/BackBone-Profile","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/praeclarumjj3%2FBackBone-Profile","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/praeclarumjj3%2FBackBone-Profile/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/praeclarumjj3%2FBackBone-Profile/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/praeclarumjj3%2FBackBone-Profile/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/praeclarumjj3","download_url":"https://codeload.github.com/praeclarumjj3/BackBone-Profile/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/praeclarumjj3%2FBackBone-Profile/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33125184,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-16T18:38:32.183Z","status":"online","status_checked_at":"2026-05-17T02:00:05.366Z","response_time":107,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["inference-performance","pytorch"],"created_at":"2024-11-17T14:51:49.216Z","updated_at":"2026-05-17T02:05:15.627Z","avatar_url":"https://github.com/praeclarumjj3.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# ResNet Profile\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)\n[![Framework: PyTorch](https://img.shields.io/badge/Framework-PyTorch-orange.svg)](https://pytorch.org/)\n\n## Contents\n1. [Overview](#1-overview)\n2. [Setup Instructions](#2-setup-instructions)\n3. [Repository Overview](#3-repository-overview)\n4. [Reproduction](#5-reproduction)\n5. [Results](#5-results)\n\n## 1. Overview\n\nThis repo contains the folowing **Performance Stats** for a few popularly used **backbone networks** in the field of Computer Vision:\n\n```\n- Inference Time on a GTX 2080Ti\n\n- Inference Time on a TitanXP\n\n- Infernce Time on a CPU\n\n- Memory Report during Inference\n\n- Model Structures\n```\n\nI have performed experiments on **two** types of inputs:\n\n`Size Format: (B,C,H,W)`\n\n- **Cityscapes:** Input of size = (1,3,1024,2048)\n\n- **PASCAL-VOC-2012:** Input of size = (1,3,500,334)\n\n## 2. Setup Instructions\n\nYou can setup the repo by running the following commands:\n```\n$ git clone https://github.com/praeclarumjj3/BackBone-Profile.git\n```\n\n```\n$ pip install -r requirements.txt\n```\n\n## 3. Repository Overview\n\nThe repository contains the following architecture:\n\n- [MobileNetV2 Profiler](https://github.com/praeclarumjj3/BackBone-Profile/tree/master/MobilnetV2%20Profiler) - Scripts and stats for inference performance of **MobileNet-V2**.\n\n- [ResNet Profiler](https://github.com/praeclarumjj3/BackBone-Profile/tree/master/ResNet%20Profiler) - Scripts and stats for inference performance of various variants of **ResNet**.\n\n- [Xception Profiler](https://github.com/praeclarumjj3/BackBone-Profile/tree/master/Xception%20Profiler) - Scripts and stats for inference performance of **Xception**.\n\n## 4. Reproduction\n\n- Refer to the **README.md** of the corresponding architectures.\n\n## 5. Results\n\nAll the experiments are performed with a `batch size=1` and 300 iterations.\n\n#### Performance on Cityscapes\n\n|     Model         | Inference Time (ms) [2080Ti] | Inference Time (ms) [TitanXP]|   FPS [2080Ti] | FPS [TitanXP]  |  Allocated Memory (MB)  | # Params (M)  |\n| ----------------- | ---------------------------- | ---------------------------- | -------------- | -------------- | ----------------------- | --------------|\n| **ResNet-18**     |       **18.719**             |       **23.622**             | **53.42**      | **42.33**      |   **68.69**             | **11.689**    |\n| **ResNet-34**     |       **31.779**             |       **38.588**             | **31.46**      | **25.91**      |   **108.16**            | **21.797**    |\n| **ResNet-50**     |       **61.397**             |       **82.334**             | **16.28**      | **12.14**      |   **121.73**            | **25.557**    |\n| **ResNet-101**    |       **100.426**            |       **122.491**            | **9.95**       | **8.16**       |   **194.65**            | **44.549**    |\n| **MobileNet-V2**  |       **33.627**             |       **54.314**             | **29.73**      | **18.41**      |   **37.58**             | **3.504**     |\n| **Xception**      |       **77.079**             |       **144.919**            | **12.97**      | **6.90**       |   **111.45**            | **22.855**     |\n\n#### Performance on PASCAL-VOC-2012\n\n|     Model         | Inference Time (ms) [2080Ti] | Inference Time (ms) [TitanXP]|   FPS [2080Ti] | FPS [TitanXP]  |  Allocated Memory (MB)  | # Params (M)  | \n| ----------------- | ---------------------------- | ---------------------------- | -------------- | -------------- | ----------------------- | --------------|\n| **ResNet-18**     |       **2.547**              |       **2.940**              | **392.61**     | **340.13**     |   **46.60**             | **11.689**    |\n| **ResNet-34**     |       **5.197**              |       **4.959**              | **192.41**     | **201.65**     |   **85.20**             | **21.797**    |\n| **ResNet-50**     |       **7.628**              |       **8.927**              | **131.09**     | **112.01**     |   **100.23**            | **25.557**    |\n| **ResNet-101**    |       **12.579**             |       **14.509**             | **79.49**      | **68.92**      |   **172.65**            | **44.549**    |\n| **MobileNet-V2**  |       **5.570**              |       **5.795**              | **179.53**     | **172.56**     |   **14.49**             | **3.504**     |\n| **Xception**      |       **7.919**              |       **12.042**             | **126.27**     | **83.04**      |   **89.36**             | **22.855**     |\n\n### Inference Time (CPU) \n\n#### Performance on Cityscapes\n\n|     Model         | Inference Time (ms) |    FPS    | \n| ----------------- | --------------------| --------- |\n| **ResNet-18**     |       **566.75**    | **1.76**  |\n| **ResNet-34**     |       **807.57**    | **1.23**  |\n| **ResNet-50**     |       **1626.05**   | **0.61**  |\n| **ResNet-101**    |       **2344.98**   | **0.42**  |\n| **MobileNet-V2**  |       **560.022**   | **1.78**  |\n| **Xception**      |       **2782.874**  | **0.35**  |\n\n#### Performance on PASCAL-VOC-2012\n\n|        Model          |    Inference Time (ms)    |       FPS        | \n|    -----------------  |    --------------------   |    ---------     |\n|    **ResNet-18**      |          **55.79**        |    **17.92**     |\n|    **ResNet-34**      |          **78.77**        |    **12.69**     |\n|    **ResNet-50**      |          **133.71**       |    **7.47**      |\n|    **ResNet-101**     |          **223.59**       |    **4.47**      |\n|    **MobileNet-V2**   |          **70.180**       |   **14.24**      | \n|    **Xception**       |          **229.000**      |   **4.36**       |","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpraeclarumjj3%2Fbackbone-profile","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpraeclarumjj3%2Fbackbone-profile","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpraeclarumjj3%2Fbackbone-profile/lists"}