Indoor Scene Recognition with a Visual Attention-Driven Spatial Pooling Strategy
Bibliographic record
Abstract
Scene recognition is an important research topic in robotics and computer vision. Even though scene recognition is a problem that has been studied in depth, indoor scene categorization has had a slow progress. Indoor scene recognition is a challenging problem due to the severe high intra-class variability, mainly due to the intrinsic variety of objects that may be present, and inter-class similarities of man-made indoor structures. Therefore, most scene recognition techniques that work well for outdoor scenes demonstrate low performance on indoor scenes. Thus, in this paper, we present a simple, yet effective method for indoor scene recognition. Our approach can be illustrated as follows. First, we extract dense SIFT descriptors. Then, we combine a saliency-driven perceptual pooling with a simple spatial pooling scheme. Once the spatial and the saliency-driven encoding have been determined, we use vector quantization to compute histograms of local features from each sub-region. Later, the histograms from all sub-regions are concatenated together to generate the final representation of the image. Finally, a model based mixture classifier, which uses mixture models to characterize class densities, is applied. In order to address the problem of modeling non-Gaussian data which are largely present in our final representation of images, we use the generalized Gaussian mixture (GGM) which can be a good alternative to the Gaussian thanks to its shape flexibility. The learning of the proposed statistical model is carried out using the rival penalized expectation-maximization (RPEM) algorithm which is able to perform model selection and parameter learning together in a single step. Furthermore, we take into account the feature selection problem by determining a set of relevant features for each data cluster, so that we can speed up the used learning algorithm and get rid of noisy, redundant, or uninformative feature. To validate the proposed method we test it on the MIT indoor scenes data set.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".