Joint-DetNAS: Upgrade Your Detector with NAS, Pruning and Dynamic Distillation
We propose Joint-DetNAS, a unified NAS framework for object detection, which integrates 3 key components: Neural Architecture Search, pruning, and Knowledge Distillation. Instead of naively pipelining these techniques, our Joint-DetNAS optimizes them jointly. The algorithm consists of two core processes: student morphism optimizes the student's architecture and removes the redundant parameters, while dynamic distillation aims to find the optimal matching teacher. For student morphism, weight inheritance strategy is adopted, allowing the student to flexibly update its architecture while fully utilize the predecessor's weights, which considerably accelerates the search; To facilitate dynamic distillation, an elastic teacher pool is trained via integrated progressive shrinking strategy, from which teacher detectors can be sampled without additional cost in subsequent searches. Given a base detector as the input, our algorithm directly outputs the derived student detector with high performance without additional training. Experiments demonstrate that our Joint-DetNAS outperforms the naive pipelining approach by a great margin. Given a classic R101-FPN as the base detector, Joint-DetNAS is able to boost its mAP from 41.4 to 43.9 on MS COCO and reduce the latency by 47%, which is on par with the SOTA EfficientDet while requiring less search cost. We hope our proposed method can provide the community with a new way of jointly optimizing NAS, KD and pruning.
Code (0)
등록된 구현이 없습니다.
Tasks
Knowledge DistillationNeural Architecture Searchobject-detectionObject DetectionMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
DetNAS: Backbone Search for Object Detection
Object detectors are usually equipped with backbone networks designed for image classification. It might be sub-optimal because of the gap between the tasks of image classification and object detection. In this work, we …
General Classificationimage-classificationImage ClassificationNeural Architecture Search+3Using machine learning to speed up new and upgrade detector studies: a calorimeter case
In this paper, we discuss the way advanced machine learning techniques allow physicists to perform in-depth studies of the realistic operating modes of the detectors during the stage of their design. Proposed approach ca…
BIG-bench Machine LearningVisual Prompting Upgrades Neural Network Sparsification: A Data-Model Perspective
The rapid development of large-scale deep learning models questions the affordability of hardware platforms, which necessitates the pruning to reduce their computational and memory footprints. Sparse neural networks as t…
Image ClassificationVisual PromptingSWA Object Detection
Do you want to improve 1.0 AP for your object detector without any inference cost and any change to your detector? Let us tell you such a recipe. It is surprisingly simple: train your detector for an extra 12 epochs usin…
Instance SegmentationObjectobject-detectionObject Detection+1Task Specific Pruning with LLM-Sieve: How Many Parameters Does Your Task Really Need?
As Large Language Models (LLMs) are increasingly being adopted for narrow tasks - such as medical question answering or sentiment analysis - and deployed in resource-constrained settings, a key question arises: how many …
Medical Question AnsweringQuantizationQuestion AnsweringSentiment Analysis