Added readme + model

Author: NotSimone

README.md

@@ -0,0 +1,9 @@
+# foo_cnn_bpm
+
+Foobar2000 plugin to analyse bpm for tracks. Based on the work of hendrik73's [tempo-cnn](https://github.com/hendriks73/tempo-cnn) and using one of his models.
+
+![screenshot](screenshot.png)
+
+Currently this will attempt to make a series of BPM predictions but it cannot write any BPM information to tags (yet).
+
+Requires a copy of the tensorflow c api v1.x dll to also be located within the foobar install directory.

model/deepsquare-k16-3.json

@@ -0,0 +1,54 @@
+{
+    "name": "DeepSquare",
+    "type": "tempo estimation",
+    "link": "https://essentia.upf.edu/models/tempo/tempocnn/deepsquare-k16-3.pb",
+    "version": "1",
+    "description": "This model outputs a softmax distribution of length 256 where the first index represents the probability of the track being 30 BPM and the last one the probability of being 286 BPM.",
+    "author": "Pablo Alonso",
+    "email": "[email protected]",
+    "release_date": "04/08/2020",
+    "framework": "tensorflow",
+    "framework_version": "2.3.0",
+    "programming_env": {
+        "language": "Python",
+        "version": "3.8"
+    },
+    "model_types": [
+        "frozen_model"
+    ],
+    "dataset": {
+        "name": "Extended Ballroom, LMDTempo and MTGTempo",
+        "size": "3826 30s excerpts, 3611 30s excerpts and 1158 2m excerpts",
+        "link": [
+            "http://anasynth.ircam.fr/home/media/ExtendedBallroom",
+            "https://github.com/GiantSteps/giantsteps-tempo-dataset"
+        ]
+    },
+    "schema": {
+        "inputs": [
+            {
+                "name": "input",
+                "type": "float",
+                "shape": [
+                    256,
+                    40
+                ]
+            }
+        ],
+        "outputs": [
+            {
+                "name": "output",
+                "op": "Softmax",
+                "shape": [
+                    256
+                ],
+                "output_purpose": "predictions"
+            }
+        ]
+    },
+    "citation": "@inproceedings{SchreiberM19_CNNKeyTempo_SMC,\nTitle = {Musical Tempo and Key Estimation using Convolutional Neural Networks with Directional Filters},\nAuthor = {Hendrik Schreiber and Meinard M{\\\"u}ller},\nBooktitle = {Proceedings of the Sound and Music Computing Conference ({SMC})},\nPages = {47--54},\nYear = {2019},\nAddress = {M{\\'a}laga, Spain}}",
+    "inference": {
+        "sample_rate": 11025,
+        "algorithm": "TempoCNN"
+    }
+}