Evolving Ambiguous Images
Authors: Penousal Machado, Adriano Vinhas, João Correia, Aniko Ekárt
IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The experimentation was divided into three steps. First we reproduced the previous work of [Correia et al., 2013], by performing evolutionary runs with a single object detector. Next, we combined the two detectors to test the ability of our approach to evolve images that simultaneously depict both objects. Finally, we combined the detectors enforcing an overlap between the objects, hoping to achieve ambiguous images. |
| Researcher Affiliation | Academia | Penousal Machado1 and Adriano Vinhas1 and Jo ao Correia1 and Aniko Ek art2 1 CISUC, Department of Informatics Engineering University of Coimbra, Coimbra, Portugal 2Aston Lab for Intelligent Collectives Engineering (ALICE), Computer Science Aston University, Birmingham, United Kingdom |
| Pseudocode | No | The paper describes the detection algorithm in numbered steps and presents an overview diagram (Figure 2), but it does not contain a formally labeled "Pseudocode" or "Algorithm" block. |
| Open Source Code | No | The paper mentions using "ECJ, a java-based Evolutionary Computation research system" and "Open CV API" (http://opencv.org/) as tools, but it does not provide an explicit statement or link to the authors' own source code implementation for the described methodology. |
| Open Datasets | No | The paper states: "Two object detectors were trained, by building datasets of faces and flowers. This training procedure was attained using Open CV API." and "The process was similar to the one used by Correia et al. [2013] where more insight about building the datasets can be found." While it mentions datasets were built and cites a paper for insight on building them, it does not provide concrete access information (link, DOI, repository name, or specific formal citation for the dataset itself with authors/year) to these datasets for public access. |
| Dataset Splits | No | The paper discusses training and detection, and mentions "30 independent evolutionary runs", but does not specify explicit training/validation/test dataset splits (e.g., percentages, sample counts, or references to predefined splits for reproducibility). |
| Hardware Specification | No | The paper does not provide any specific details about the hardware (e.g., CPU, GPU models, memory) used to run the experiments. |
| Software Dependencies | No | The paper mentions "ECJ, a java-based Evolutionary Computation research system" and "Open CV API" but does not specify version numbers for these software components. |
| Experiment Setup | Yes | Table 1: Parameters of the GP engine. Parameter Setting Population size 100 Number of generations 1000 Crossover probability 0.8 (per individual) Mutation probability 0.05 (per node) Mutation operators sub-tree swap, sub-tree replacement, node insertion, node deletion and mutation Initialization method ramped half-and-half Initial maximum depth 5 Mutation max tree depth 3 Function set +, , , /, min, max, abs, neg, warp, sign, sqrt, pow, mdist, sin, cos, if Terminal set x, y, random constants. Table 2: Detection parameters used. Parameter Setting Min. window width 90 Min. window height 90 Image Width 128 Image Height 128 Scale Factor 1.1 Image pre-processing Otsu s Binarization |