Estimate and Replace: A Novel Approach to Integrating Deep Neural Networks with Existing Applications

2018/04/24 by Guy Hadash, Einat Kermany, Hadash, Guy +9 · 5 citations
Computer Science · #Computation and Language (cs.CL) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Multimodal Machine Learning Applications #Reinforcement Learning in Robotics #Topic Modeling

paper · pdf · doi:10.48550/arxiv.1804.09028

openalex publication_date 2018/04/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Existing applications include a huge amount of knowledge that is out of reach for deep neural networks. This paper presents a novel approach for integrating calls to existing applications into deep learning architectures. Using this approach, we estimate each application's functionality with an estimator, which is implemented as a deep neural network (DNN). The estimator is then embedded into a base network that we direct into complying with the application's interface during an end-to-end optimization process. At inference time, we replace each estimator with its existing application counterpart and let the base network solve the task by interacting with the existing application. Using this 'Estimate and Replace' method, we were able to train a DNN end-to-end with less data and outperformed a matching DNN that did not interact with the external application.

Citations

• Long Short-Term Memory
• Compositional Semantic Parsing on Semi-Structured Tables
• Neural Programmer-Interpreters
• Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
• Achieving Human Parity in Conversational Speech Recognition
• Categorical Reparameterization with Gumbel-Softmax

Cited by

• Fine-tuning Pre-trained Audio Models for COVID-19 Detection: A Technical Report
• Layer-Wise Perturbations via Sparse Autoencoders for Adversarial Text Generation
• An h-space Based Adversarial Attack for Protection Against Few-shot Personalization
• Agent-Based Detection and Resolution of Incompleteness and Ambiguity in Interactions with Large Language Models
• Noise Optimization for Artificial Neural Networks

Related

• Reusing Deep Learning Models: Challenges and Directions in Software Engineering
• Repairing DNN Architecture: Are We There Yet?
• Reusing Deep Neural Network Models through Model Re-engineering
• Tailor: Altering Skip Connections for Resource-Efficient Inference
• Deploying Deep Neural Networks in the Embedded Space