paper-with-me

Papers

End-to-End Supermask Pruning: Learning to Prune Image Captioning Models

2021-10-07 · Jia Huei Tan, Chee Seng Chan, Joon Huang Chuah

With the advancement of deep models, research work on image captioning has led to a remarkable gain in raw performance over the last decade, along with increasing model complexity and computational cost. However, surprisingly works on compression of deep networks for image captioning task has received little to no attention. For the first time in image captioning research, we provide an extensive comparison of various unstructured weight pruning methods on three different popular image captioning architectures, namely Soft-Attention, Up-Down and Object Relation Transformer. Following this, we propose a novel end-to-end weight pruning method that performs gradual sparsification based on weight sensitivity to the training loss. The pruning schemes are then extended with encoder pruning, where we show that conducting both decoder pruning and training simultaneously prior to the encoder pruning provides good overall performance. Empirically, we show that an 80% to 95% sparse network (up to 75% reduction in model size) can either match or outperform its dense counterpart. The code and pre-trained models for Up-Down and Object Relation Transformer that are capable of achieving CIDEr scores >120 on the MS-COCO dataset but with only 8.7 MB and 14.5 MB in model size (size reduction of 96% and 94% respectively against dense versions) are publicly available at https://github.com/jiahuei/sparse-image-captioning.

📄 PDF Abstract BibTeX arXiv:2110.03298

Code (1)

jiahuei/sparse-image-captioning 공식 구현 pytorch

Tasks

DecoderImage Captioning

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Pruning 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Position-Wise Feed-Forward Layer 설명 없음
Adam 설명 없음
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Label Smoothing Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…

Similar Papers 제목 키워드 기반

HideNseek: Federated Lottery Ticket via Server-side Pruning and Sign Supermask

2022-06-09 · Anish K. Vallapuram, Pengyuan Zhou, Young D. Kwon, Lik Hang Lee 외

Federated learning alleviates the privacy risk in distributed learning by transmitting only the local model updates to the central server. However, it faces challenges including statistical heterogeneity of clients' data…

Federated LearningModel CompressionQuantization

OSSuM: A Gradient-Free Approach For Pruning Neural Networks At Initialization

2021-09-29 · Vinu Sankar Sadasivan, Jayesh Malaviya, Anirban Dasgupta

Pruning overparameterized neural networks to obtain memory-and-compute-efficient sparse networks is an active area of research. Recent works attempt to prune neural networks at initialization to design sparse networks th…

Audio-Visual Exchange-Aware Token Pruning for Efficient Audio-Visual Captioning

2026-06-09 · Zihan Meng, Dexiang Hong, Weidong Chen, Ziyu Zhou 외 arxiv

Audio-visual captioning generates natural language descriptions from video and audio content. Multimodal LLMs have advanced this task, but both modalities contribute many tokens to the LLM input, where prefill self-atten…

Data-Free Backbone Fine-Tuning for Pruned Neural Networks

2023-06-22 · Adrian Holzbock, Achyut Hegde, Klaus Dietmayer, Vasileios Belagiannis

Model compression techniques reduce the computational load and memory consumption of deep neural networks. After the compression operation, e.g. parameter pruning, the model is normally fine-tuned on the original trainin…

2D Human Pose Estimationimage-classificationImage ClassificationModel Compression+3

EvoPrune: Early-Stage Visual Token Pruning for Efficient MLLMs

2026-03-04 · Yuhao Chen, Bin Shan, Xin Ye, Cheng Chen arxiv

Multimodal Large Language Models (MLLMs) have shown strong performance in vision-language tasks, but their inference efficiency is severely limited by the exponential growth of visual tokens in complex scenarios such as …