Non-parametric Structured Output Networks

Authors: Andreas Lehrmann, Leonid Sigal

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We validate non-parametric structured output networks at both the model (DNN+NGM) and the inference level (RIN). Model validation consists of a comparison to baselines along two binary axes, structuredness and non-parametricity. Inference validation compares our RIN unit to the two predominant groups of approaches for inference in structured non-parametric densities, i.e., sampling-based and variational inference (Section 1.1.2). [...] Table 2: Quantitative Evaluation.
Researcher Affiliation Industry Andreas M. Lehrmann Disney Research Pittsburgh, PA 15213 andreas.lehrmann@disneyresearch.com Leonid Sigal Disney Research Pittsburgh, PA 15213 lsigal@disneyresearch.com
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement or link indicating that the source code for the methodology is openly available.
Open Datasets Yes We test our approach on simple natural pixel statistics from Microsoft COCO [11] by sampling stripes y = (yi)n i=1 2 [0, 255]n of n = 10 pixels. [11] Lin, T.Y., Maire, M., Belongie, S., Bourdev, L., Girshick, R., Hays, J., Perona, P., Ramanan, D., Zitnick, C.L., Dollar, P.: Microsoft COCO: Common Objects in Context. In ar Xiv:1405.0312 [cs.CV]. (2014)
Dataset Splits No Using this noise process, we generate training and test sets of sizes 100,000 and 1,000, respectively. The paper does not explicitly specify a validation split.
Hardware Specification No The paper does not provide specific hardware details (e.g., exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment.
Experiment Setup Yes We train the other 4 models for 40 epochs using a Gaussian kernel and a diagonal bandwidth matrix for the non-parametric models. The DNN consists of 2 fully-connected layers with 256 units and the kernel weights are constrained to lie on a simplex with a softmax layer. The NGM uses a chain-structured graph that connects each pixel to its immediate neighbors.