Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Research on color recipe recommendation based on unstructured data using TENN

Seongsu Jhang, Donghwi Yoo, Jaeyong Kown

TL;DR

This work tackles the digitization of tacit color engineering knowledge in SMEs by inferring color recipes from unstructured, emotion-rich natural language. It introduces TENN, a Tokenizing Encoder Neural Network, trained on an RGB color-code dataset where textual descriptions map to RGB triplets. Evaluation on a dataset of roughly 300 samples shows that MinMaxScaler preprocessing yields the best accuracy around 0.79, while color similarity measured by $\Delta E$ under $CIEDE2000$ averages 73.8 across 30 test cases, indicating room for improvement but feasibility. The study demonstrates the potential of emotion-aware, unstructured-data color inference and outlines future directions including real pigment datasets, PoC collaborations with SMEs, and privacy-preserving techniques such as federated learning.

Abstract

Recently, services and business models based on large language models, such as OpenAI Chatgpt, Google BARD, and Microsoft copilot, have been introduced, and the applications utilizing natural language processing with deep learning are increasing, and it is one of the natural language preprocessing methods. Conversion to machine language through tokenization and processing of unstructured data are increasing. Although algorithms that can understand and apply human language are becoming increasingly sophisticated, it is difficult to apply them to processes that rely on human emotions and senses in industries that still mainly deal with standardized data. In particular, in processes where brightness, saturation, and color information are essential, such as painting and injection molding, most small and medium-sized companies, excluding large corporations, rely on the tacit knowledge and sensibility of color mixers, and even customer companies often present non-standardized requirements. . In this paper, we proposed TENN to infer color recipe based on unstructured data with emotional natural language, and demonstrated it.

Research on color recipe recommendation based on unstructured data using TENN

TL;DR

This work tackles the digitization of tacit color engineering knowledge in SMEs by inferring color recipes from unstructured, emotion-rich natural language. It introduces TENN, a Tokenizing Encoder Neural Network, trained on an RGB color-code dataset where textual descriptions map to RGB triplets. Evaluation on a dataset of roughly 300 samples shows that MinMaxScaler preprocessing yields the best accuracy around 0.79, while color similarity measured by under averages 73.8 across 30 test cases, indicating room for improvement but feasibility. The study demonstrates the potential of emotion-aware, unstructured-data color inference and outlines future directions including real pigment datasets, PoC collaborations with SMEs, and privacy-preserving techniques such as federated learning.

Abstract

Recently, services and business models based on large language models, such as OpenAI Chatgpt, Google BARD, and Microsoft copilot, have been introduced, and the applications utilizing natural language processing with deep learning are increasing, and it is one of the natural language preprocessing methods. Conversion to machine language through tokenization and processing of unstructured data are increasing. Although algorithms that can understand and apply human language are becoming increasingly sophisticated, it is difficult to apply them to processes that rely on human emotions and senses in industries that still mainly deal with standardized data. In particular, in processes where brightness, saturation, and color information are essential, such as painting and injection molding, most small and medium-sized companies, excluding large corporations, rely on the tacit knowledge and sensibility of color mixers, and even customer companies often present non-standardized requirements. . In this paper, we proposed TENN to infer color recipe based on unstructured data with emotional natural language, and demonstrated it.
Paper Structure (5 sections, 4 figures, 1 table)

This paper contains 5 sections, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Dataset and Model
  3. Experiments
  4. Results
  5. Conclusion

Figures (4)

  • Figure 1: Model architecture
  • Figure 2: Data pre-processing, RGB color recipe
  • Figure 3: Model architecture
  • Figure 4: Model architecture