Maximum-Entropy Fine Grained Classification

Authors: Abhimanyu Dubey, Otkrist Gupta, Ramesh Raskar, Nikhil Naik

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We provide a theoretical as well as empirical justification of our approach, and achieve state-of-the-art performance across a variety of classification tasks in FGVC
Researcher Affiliation Academia Abhimanyu Dubey Otkrist Gupta Ramesh Raskar Nikhil Naik Massachusetts Institute of Technology Cambridge, MA, USA {dubeya, otkrist, raskar, naik}@mit.edu
Pseudocode No The paper describes the objective function and theoretical formulations, but does not include any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not contain any explicit statements about releasing source code for their methodology or provide links to a code repository.
Open Datasets Yes We obtain state-of-the-art results on all five datasets (Table 1-(A-E))... CUB-200-2011 [44], Cars [22], Aircrafts [28], NABirds [43], Stanford Dogs [19]... We evaluate Maximum-Entropy on the CIFAR-10 and CIFAR-100 datasets [23]... Image Net [7]
Dataset Splits No The paper mentions 'validation sets' (e.g., 'Image Net validation set') and '5-way cross-validation' for one specific experiment, but does not provide explicit overall train/validation/test split percentages or sample counts for the main FGVC datasets.
Hardware Specification Yes We perform all experiments using the Py Torch [32] framework over a cluster of NVIDIA Titan X GPUs.
Software Dependencies No The paper mentions 'Py Torch [32]' as the framework used but does not provide a specific version number for it or any other software dependencies.
Experiment Setup No The paper discusses the robustness to the choice of hyperparameter γ and label noise, but it does not provide specific details such as learning rate, batch size, number of epochs, or optimizer settings for the experimental setup.