High Performance Large Scale Image Recognition Without Normalization

Author(s)	Priyanka Sharma, Pratima Sharma
Country	India
Abstract	ABSTRACT Load standardization is an important aspect of most models, but the usage is focused on a varied range of undesirable features as it depends on the load size and affects the interactions of instances. However, newest profound resent networks are capable of being trained without any standardizing layer, but their precise operation is not the same as standardized batch networks. This will create an adaptive gradient cutting system that fixes these instabilities and develops a much improved Free ResNet class. Our lesser edition is 8.7% faster, and our largest is 87% faster than the top 1.0. The Net-B7 is a versatile Picture Net measuring precision for our smaller versions. As demonstrated in our top models with the accuracy of approximately 90%, Free Normalize models significantly increases the efficiency in fine tuning, compared with large-scale pre-training with 350 million pictures of data-sets.
Keywords	Keywords: Image Recognition, Normalization, ResNet, Tuning, Large-scale Pre-training
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 6, Issue 4, July-August 2024
Published On	2024-08-31
Cite This	High Performance Large Scale Image Recognition Without Normalization - Priyanka Sharma, Pratima Sharma - IJFMR Volume 6, Issue 4, July-August 2024. DOI 10.36948/ijfmr.2024.v06i04.26656
DOI	https://doi.org/10.36948/ijfmr.2024.v06i04.26656
Short DOI	https://doi.org/gt9hds

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9898-5948-55	editor@ijfmr.com

International Journal For Multidisciplinary Research