SubmodBoxes: Near-Optimal Search for a Set of Diverse Object Proposals
Authors: Qing Sun, Dhruv Batra
NeurIPS 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirically, we show that our approach leads to a state-of-art performance on object proposal generation via a novel diversity measure.We apply our proposed technique Submod Boxes to the task of generating object proposals [2,7,39, 41] on the PASCAL VOC 2007 [13], PASCAL VOC 2012 [14], and MS COCO [28] datasets. Our results show that our approach outperforms all baselines. |
| Researcher Affiliation | Academia | Qing Sun Virginia Tech sunqing@vt.edu Dhruv Batra Virginia Tech https://mlp.ece.vt.edu/ |
| Pseudocode | No | The paper describes algorithms procedurally and with diagrams, but does not include a formally labeled 'Pseudocode' or 'Algorithm' block. |
| Open Source Code | No | The paper does not provide any statement or link indicating the availability of source code for the described methodology. |
| Open Datasets | Yes | We evaluate Submod Boxes for object proposal generation on three datasets: PASCAL VOC 2007 [13], PASCAL VOC 2012 [14], and MS COCO [28]. |
| Dataset Splits | Yes | For the two PASCAL datasets, we perform cross validation on 2510 validation images of PASCAL VOC 2007 for the best parameter λ, then report accuracies on 4952 test images of PASCAL VOC 2007 and 5823 validation images of PASCAL VOC 2012. The MS COCO dataset is much larger, so we randomly select a subset of 5000 training images for tuning λ, and test on complete validation dataset with 40138 images. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory, or specific machine types) used for running the experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details, such as library names with version numbers. |
| Experiment Setup | Yes | In a manner similar to [23], we chose a different λM for M = 100, 200, 400, 600, 800, 1000 proposals.At λ = 0 all methods produce the same (highest weighted) box M times. At λ = , they all perform a reranking of the reference set of boxes.Instead of allowing the chosen boxes to cover exactly one reference box, we analyze the effect of matching top-k reference boxes. |