使用 amc 进行异常值检测matlab代码.zip

# Outlier Detection using Saliency Detection via Absorbing Markov Chains Course Project for CS5150 - Applications of Markov Chain in Computing **Work by:** \ Shreya Kumar - ES20BTECH11026 \ Vikhyath Kothamasu - CS20BTECH11056 \ U Adithyan - CS23BTNCL11001 ## How to run the code (i) Generating the Heatmaps - Run the command `python density_map.py --dataset={dataset_name}` where `dataset_name` can be any one of the four values: iris, diabetes, wine, breast_cancer. This should generate 2 files: `{dataset_name}_density.png` and `{dataset_name}_metadata.json`. (ii) Generating Boundary Map: - Change the input and output path as required in `edges-master/edgesDemo.m` and run the file. - The input file would be the density map image generated from the previous step. - Ensure that the boundary map is stored under `boundary-map/` with the same name as the input image. (iii) Generating FCN features - Run the command `python fcn.py --dataset={dataset_name}` where `dataset_name` is same as the one used in first step. This should generate 2 files: `{dataset_name}_l6.mat` and `{dataset_name}_l32.mat`. (iv) Generating Saliency Map - Store the input images (.png format) in `image/` - Store their corresponding boundary maps in `boundary-map/` - Store their FCN-features in `FCN-feature/6/` and `FCN-feature/32/`. - Run `Saliency_AMC_AE.m` (v) Making Final Inferences - Run the command `python inference.py --dataset={dataset_name}` where `dataset_name` is same as the one used in first step. This should print final results and save an image containing all the scatter plots. ## References: [1] B. Jiang, L. Zhang, H. Lu, C. Yang and M. -H. Yang, "Saliency Detection via Absorbing Markov Chain," 2013 IEEE International Conference on Computer Vision, Sydney, NSW, Australia, 2013, pp. 1665-1672, doi: 10.1109/ICCV.2013.209. [2] Zitnick, C.L., Dollár, P. (2014). Edge Boxes: Locating Object Proposals from Edges. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds) Computer Vision – ECCV 2014. ECCV 2014. Lecture Notes in Computer Science, vol 8693. Springer, Cham. https://doi.org/10.1007/978-3-319-10602-1_26 **Note:** 1. The saliency maps were generated using [1] 2. Boundary maps required for [1] were generated using [2]