This README.txt file was generated on <20200304> by Nicholas Blauch # # General instructions for completing README: # For sections that are non-applicable, mark as N/A (do not delete any sections). # Please leave all commented sections in README (do not delete any text). # ------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset: Computational insights into human perceptual expertise for familiar and unfamiliar face recognition This dataset is associated with the paper at https://psyarxiv.com/bv5mp # # Authors: Include contact information for at least the # first author and corresponding author (if not the same), # specifically email address, phone number (optional, but preferred), and institution. # Contact information for all authors is preferred. # 2. Author Information First/Corresponding Author Contact Information Name: Nicholas Blauch Institution: Carnegie Mellon University Address: Baker Hall 342C, 4825 Frew St, Pittsburgh PA 15213 United States Email: blauch@cmu.edu Phone Number: n/a --------------------- DATA & FILE OVERVIEW --------------------- # # Directory of Files in Dataset: List and define the different # files included in the dataset. This serves as its table of # contents. # Directory of Files: etc/ vggface2_ids_in_vggface.txt -> a list of overlapping identities between vggface2 and vggface face_matching/ (all the data relevant to the human behavioral experiment, including raw, intermediate, and post-processed results) dcnn/ (files generated by layerwise DCNN analyses of images shown to subjects) full_verification/ (files generated by the cognitive model using perceptual and identity representations of the DCNN, on images shown to subjects) processed/ (intermediate files for processing subject data) ratings_exp-fam1_sub-*.pkl (Raw output files for each subject) facebyface/ (data for the familiarization experiments using a variable number of fine-tuned identities, corresponding to section 3.7 of the paper) models/ (model state dicts) results/ (result files during training) fine_tuning/ (data for the familarization experiments using a constant number of fine-tuned identities, corresponding to section 3.2-3.6 of the paper) models/ (model state dicts) results/ (result files during training) from_scratch/ (data for from-scratch training of DCNNs) models/ (model state dicts) results/ (result files during training) imagesets/ (imagesets used for fine-tuning experiments) Additional Notes on File Relationships, Context, or Content (for example, if a user wants to reuse and/or cite your data, what information would you want them to know?): There is an associated code repository at https://github.com/familiarity_sims . Questions should generally be posted there (as "issues"), and users may find looking through issues helpful for debugging their own problems. # # File Naming Convention: Define your File Naming Convention # (FNC), the framework used for naming your files systematically # to describe what they contain, which could be combined with the # Directory of Files. # File Naming Convention: Generally, data files are named to include some set of key/value pairs, e.g.: key1-val1_key2-val2.ext Image sets are stored in a conventional format, as: ///. # # Data Description: A data description, dictionary, or codebook # defines the variables and abbreviations used in a dataset. This # information can be included in the README file, in a separate # file, or as part of the data file. If it is in a separate file # or in the data file, explain where this information is located # and ensure that it is accessible without specialized software. # (We recommend using plain text files or tabular plain text CSV # files exported from spreadsheet software.) # ----------------------------------------- DATA DESCRIPTION FOR: N/A ----------------------------------------- 1. Number of variables: 2. Number of cases/rows: 3. Missing data codes: Code/symbol Definition Code/symbol Definition 4. Variable List # # Example. Name: Gender # Description: Gender of respondent # 1 = Male # 2 = Female # 3 = Transgender # 4 = Nonbinary # 5 = Other gender not listed # 6 = Prefer not to answer # A. Name: Description: Value labels if appropriate B. Name: Description: Value labels if appropriate -------------------------- METHODOLOGICAL INFORMATION -------------------------- # # Software: If specialized software(s) generated your data or # are necessary to interpret it, please provide for each (if # applicable): software name, version, system requirements, # and developer. #If you developed the software, please provide (if applicable): #A copy of the software’s binary executable compatible with the system requirements described above. #A source snapshot or distribution if the source code is not stored in a publicly available online repository. #All software source components, including pointers to source(s) for third-party components (if any) 1. Software-specific information: Name: Version: System Requirements: Open Source? (Y/N): (if available and applicable) Executable URL: Source Repository URL: Developer: Product URL: Software source components: Additional Notes(such as, will this software not run on certain operating systems?): See https://github.com/familiarity_sims for instructions on creating a Python environment that can accurately reproduce these data. # # Equipment: If specialized equipment generated your data, # please provide for each (if applicable): equipment name, # manufacturer, model, and calibration information. Be sure # to include specialized file format information in the data # dictionary. # 2. Equipment-specific information: N/A Manufacturer: Model: (if applicable) Embedded Software / Firmware Name: Embedded Software / Firmware Version: Additional Notes: # # Dates of Data Collection: List the dates and/or times of # data collection. # 3. Date of data collection (single date, range, approximate date) : January 2019-February 2020