Allow to run custom classifier files without the labelsfile (#719)

* Allow to run custom classifier files without the labelsfile

* added i8n

Author: Josef-Haupt

birdnet_analyzer/analyze/core.py

@@ -207,7 +207,11 @@ def _set_params(
             if not os.path.isfile(cfg.LABELS_FILE):
                 cfg.LABELS_FILE = custom_classifier.replace("Model_FP32.tflite", "Labels.txt")
 
-            cfg.LABELS = read_lines(cfg.LABELS_FILE)
+            if not custom_classifier.endswith("Model_FP32.tflite") or not os.path.isfile(cfg.LABELS_FILE):
+                cfg.LABELS_FILE = None
+                cfg.LABELS = None
+            else:
+                cfg.LABELS = read_lines(cfg.LABELS_FILE)
         else:
             cfg.APPLY_SIGMOID = False
             # our output format
@@ -236,10 +240,13 @@ def _set_params(
             cfg.SPECIES_LIST_FILE = None
             cfg.SPECIES_LIST = get_species_list(cfg.LATITUDE, cfg.LONGITUDE, cfg.WEEK, cfg.LOCATION_FILTER_THRESHOLD)
 
-    lfile = os.path.join(cfg.TRANSLATED_LABELS_PATH, os.path.basename(cfg.LABELS_FILE).replace(".txt", f"_{locale}.txt"))
+    if cfg.LABELS_FILE:
+        lfile = os.path.join(cfg.TRANSLATED_LABELS_PATH, os.path.basename(cfg.LABELS_FILE).replace(".txt", f"_{locale}.txt"))
 
-    if locale not in ["en"] and os.path.isfile(lfile):
-        cfg.TRANSLATED_LABELS = read_lines(lfile)
+        if locale not in ["en"] and os.path.isfile(lfile):
+            cfg.TRANSLATED_LABELS = read_lines(lfile)
+        else:
+            cfg.TRANSLATED_LABELS = cfg.LABELS
     else:
         cfg.TRANSLATED_LABELS = cfg.LABELS
 

birdnet_analyzer/analyze/utils.py

@@ -90,7 +90,7 @@ def generate_raven_table(timestamps: list[str], result: dict[str, list], afile_p
 
         for c in result[timestamp]:
             selection_id += 1
-            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])]
+            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])] if cfg.TRANSLATED_LABELS else c[0]
             code = cfg.CODES[c[0]] if c[0] in cfg.CODES else c[0]
             rstring += (
                 f"{selection_id}\tSpectrogram 1\t1\t{start}\t{end}\t{low_freq}\t{high_freq}\t{label.split('_', 1)[-1]}\t{code}\t{c[1]:.4f}\t{afile_path}\t{start}\n"
@@ -130,7 +130,7 @@ def generate_audacity(timestamps: list[str], result: dict[str, list], result_pat
         rstring = ""
 
         for c in result[timestamp]:
-            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])]
+            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])] if cfg.TRANSLATED_LABELS else c[0]
             ts = timestamp.replace("-", "\t")
             lbl = label.replace("_", ", ")
             rstring += f"{ts}\t{lbl}\t{c[1]:.4f}\n"
@@ -165,7 +165,7 @@ def generate_kaleidoscope(timestamps: list[str], result: dict[str, list], afile_
         start, end = timestamp.split("-", 1)
 
         for c in result[timestamp]:
-            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])]
+            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])] if cfg.TRANSLATED_LABELS else c[0]
             rstring += "{},{},{},{},{},{},{},{:.4f},{:.4f},{:.4f},{},{},{}\n".format(
                 parent_folder.rstrip("/"),
                 folder_name,
@@ -220,7 +220,7 @@ def generate_csv(timestamps: list[str], result: dict[str, list], afile_path: str
 
         for c in result[timestamp]:
             start, end = timestamp.split("-", 1)
-            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])]
+            label = cfg.TRANSLATED_LABELS[cfg.LABELS.index(c[0])] if cfg.TRANSLATED_LABELS else c[0]
             rstring += f"{start},{end},{label.split('_', 1)[0]},{label.split('_', 1)[-1]},{c[1]:.4f},{afile_path}"
 
             if columns_map:
@@ -648,6 +648,9 @@ def analyze_file(item) -> dict[str, str] | None:
                     # Get prediction
                     pred = p[i]
 
+                    if not cfg.LABELS:
+                        cfg.LABELS = [f"Species-{i}_Species-{i}" for i in range(len(pred))]
+
                     # Assign scores to labels
                     p_labels = [
                         p

birdnet_analyzer/gui/utils.py

@@ -596,15 +596,21 @@ def species_lists(opened=True):
             def on_custom_classifier_selection_click():
                 file = select_file(("TFLite classifier (*.tflite)",), state_key="custom_classifier_file")
 
-                if file:
-                    labels = os.path.splitext(file)[0] + "_Labels.txt"
+                if not file:
+                    return None, None
 
-                    if not os.path.isfile(labels):
-                        labels = file.replace("Model_FP32.tflite", "Labels.txt")
+                base_name = os.path.splitext(file)[0]
+                labels = base_name + "_Labels.txt"
 
-                    return file, gr.File(value=[file, labels], visible=True)
+                if not os.path.isfile(labels):
+                    labels = file.replace("Model_FP32.tflite", "Labels.txt")
 
-                return None, None
+                    if not file.endswith("Model_FP32.tflite") or not os.path.isfile(labels):
+                        gr.Warning(loc.localize("species-list-custom-classifier-no-labelfile-warning"))
+
+                        return file, gr.File(value=[file], visible=True)
+
+                return file, gr.File(value=[file, labels], visible=True)
 
             classifier_selection_button.click(
                 on_custom_classifier_selection_click,

birdnet_analyzer/lang/de.json

@@ -2,6 +2,7 @@
     "analyze-locale-dropdown-info": "Lokalisierung für die übersetzten Trivialnamen der Arten in der Ausgabe",
     "analyze-locale-dropdown-label": "Lokalisierung",
     "analyze-start-button-label": "Analysieren",
+    "embedding-file-output-accordion-label": "Dateiausgabe",
     "embedding-settings-accordion-label": "Einstellungen",
     "embedding-settings-audio-speed-slider-info": "Negative Werte verlangsamen das Audio, positive Werte beschleunigen es. Die Einstellung wird in der Datenbank gespeichert.",
     "embedding-settings-audio-speed-slider-label": "Audio Geschwindigkeitsanpassung",
@@ -18,10 +19,6 @@
     "embeddings-db-already-exists-info": "Die Datenbank ist bereits vorhanden, wobei die Optionen für Audiogeschwindigkeit und Bandpassfilter aus den Datenbankeinstellungen verwendet werden.",
     "embeddings-db-dir-validation-message": "Bitte wählen Sie ein Datenbankverzeichnis aus",
     "embeddings-db-name-validation-message": "Bitte geben Sie einen Namen für die Datenbank an",
-    "embedding-file-output-accordion-label": "Dateiausgabe",
-    "embeddings-tab-file-output-directory-textbox-placeholder": "Wenn nicht ausgewählt, werden keine Ausgabedateien erzeugt.",
-    "embeddings-tab-file-output-directory-textbox-label": "Ausgabeverzeichnis",
-    "embeddings-select-file-output-directory-button-label": "Ausgabeverzeichnis auswählen",
     "embeddings-extract-tab-title": "Extrahieren",
     "embeddings-input-dir-validation-message": "Bitte wählen Sie ein Eingabeverzeichnis aus",
     "embeddings-search-crop-mode-radio-info": "Anpassen, wie das Abfrage Beispiel zugeschnitten wird, falls es länger als der Modellinput ist.",
@@ -45,9 +42,12 @@
     "embeddings-search-select-query-button-label": "Abfrage Beispiel auswählen",
     "embeddings-search-start-button-label": "Suche starten",
     "embeddings-search-tab-title": "Suche",
+    "embeddings-select-file-output-directory-button-label": "Ausgabeverzeichnis auswählen",
     "embeddings-tab-dataset-info": "Wählen Sie den Namen für das Dataset",
     "embeddings-tab-dataset-label": "Dataset",
     "embeddings-tab-db-info": "Der Name der Datenbank",
+    "embeddings-tab-file-output-directory-textbox-label": "Ausgabeverzeichnis",
+    "embeddings-tab-file-output-directory-textbox-placeholder": "Wenn nicht ausgewählt, werden keine Ausgabedateien erzeugt.",
     "embeddings-tab-finish-info": "Extrahieren der Embeddings abgeschlossen. Gespeichert nach:",
     "embeddings-tab-select-db-directory-button-label": "Wählen Sie das Datenbankverzeichnis",
     "embeddings-tab-select-input-directory-button-label": "Wählen Sie das Audioverzeichnis (rekursiv)",
@@ -240,6 +240,7 @@
     "species-list-coordinates-week-slider-info": "Geben Sie die Woche des Jahres an, in der die Aufnahme gemacht wurde. Jeder Monat ist dabei in vier Wochen unterteilt wird. Wählen Sie einen Wert von 1 bis 48.",
     "species-list-coordinates-week-slider-label": "Woche",
     "species-list-coordinates-yearlong-checkbox-label": "Ganzjährig",
+    "species-list-custom-classifier-no-labelfile-warning": "Keine Label-Datei im Format <customclassifier>_Labels.txt gefunden.",
     "species-list-custom-classifier-selection-button-label": "Klassifikator auswählen",
     "species-list-custom-list-file-label": "Datei",
     "species-list-radio-info": "Filtern Sie Arten, die in der Ausgabe enthalten sind.",

birdnet_analyzer/lang/en.json

@@ -2,6 +2,7 @@
     "analyze-locale-dropdown-info": "Locale for the translated species common names in the output",
     "analyze-locale-dropdown-label": "Locale",
     "analyze-start-button-label": "Analyze",
+    "embedding-file-output-accordion-label": "File output settings",
     "embedding-settings-accordion-label": "Settings",
     "embedding-settings-audio-speed-slider-info": "Negative values slow down, positive values speed up. The audio speed modification will be stored in the database.",
     "embedding-settings-audio-speed-slider-label": "Audio speed modification",
@@ -18,10 +19,6 @@
     "embeddings-db-already-exists-info": "Database already exists, using audio speed and bandpass filter options from database settings.",
     "embeddings-db-dir-validation-message": "Please select a directory for the database",
     "embeddings-db-name-validation-message": "Please specify a name for the database",
-    "embedding-file-output-accordion-label": "File output settings",
-    "embeddings-tab-file-output-directory-textbox-placeholder": "If not selected, file output will be omitted.",
-    "embeddings-tab-file-output-directory-textbox-label": "File output directory",
-    "embeddings-select-file-output-directory-button-label": "Select file output directory",
     "embeddings-extract-tab-title": "Extract",
     "embeddings-input-dir-validation-message": "Please select an input directory",
     "embeddings-search-crop-mode-radio-info": "Adjust how to crop the query sample if it is longer than the model input.",
@@ -45,9 +42,12 @@
     "embeddings-search-select-query-button-label": "Select query sample",
     "embeddings-search-start-button-label": "Start search",
     "embeddings-search-tab-title": "Search",
+    "embeddings-select-file-output-directory-button-label": "Select file output directory",
     "embeddings-tab-dataset-info": "Specify the dataset name",
     "embeddings-tab-dataset-label": "Dataset",
     "embeddings-tab-db-info": "The name of the database",
+    "embeddings-tab-file-output-directory-textbox-label": "File output directory",
+    "embeddings-tab-file-output-directory-textbox-placeholder": "If not selected, file output will be omitted.",
     "embeddings-tab-finish-info": "Finished extracting embeddings to:",
     "embeddings-tab-select-db-directory-button-label": "Select database directory",
     "embeddings-tab-select-input-directory-button-label": "Select input directory (recursive)",
@@ -240,6 +240,7 @@
     "species-list-coordinates-week-slider-info": "Specify the week of the year the recording was made, using a simplified system where each month is divided into four weeks. Choose a value from 1 to 48.",
     "species-list-coordinates-week-slider-label": "Week",
     "species-list-coordinates-yearlong-checkbox-label": "Year-round",
+    "species-list-custom-classifier-no-labelfile-warning": "No label file found in the format <customclassifier>_Labels.txt.",
     "species-list-custom-classifier-selection-button-label": "Select classifier",
     "species-list-custom-list-file-label": "File",
     "species-list-radio-info": "Filter species that are included in output.",

birdnet_analyzer/lang/fi.json

@@ -2,6 +2,7 @@
     "analyze-locale-dropdown-info": "Kieli käännetyille lajien yleisnimille tuloksissa",
     "analyze-locale-dropdown-label": "Kieli",
     "analyze-start-button-label": "Analysoi",
+    "embedding-file-output-accordion-label": "Tiedostotulosteen asetukset",
     "embedding-settings-accordion-label": "Asetukset",
     "embedding-settings-audio-speed-slider-info": "Negatiiviset arvot hidastavat, positiiviset nopeuttavat. Äänen nopeuden muutos tallennetaan tietokantaan.",
     "embedding-settings-audio-speed-slider-label": "Äänen nopeuden muutos",
@@ -41,18 +42,17 @@
     "embeddings-search-select-query-button-label": "Valitse kysely-näyte",
     "embeddings-search-start-button-label": "Aloita haku",
     "embeddings-search-tab-title": "Haku",
+    "embeddings-select-file-output-directory-button-label": "Valitse tiedostotulosteen hakemisto",
     "embeddings-tab-dataset-info": "Määritä tietoaineiston nimi",
     "embeddings-tab-dataset-label": "Tietoaineisto",
     "embeddings-tab-db-info": "Tietokannan nimi",
+    "embeddings-tab-file-output-directory-textbox-label": "Tiedostotulosteen hakemisto",
+    "embeddings-tab-file-output-directory-textbox-placeholder": "Jos ei valittu, tiedostotulostetta ei luoda.",
     "embeddings-tab-finish-info": "Upotukset poimittu kohteeseen:",
     "embeddings-tab-select-db-directory-button-label": "Valitse tietokantahakemisto",
     "embeddings-tab-select-input-directory-button-label": "Valitse syötehakemisto (rekursiivinen)",
     "embeddings-tab-start-button-label": "Poimi upotukset",
     "embeddings-tab-title": "Upotukset",
-    "embedding-file-output-accordion-label": "Tiedostotulosteen asetukset",
-    "embeddings-tab-file-output-directory-textbox-placeholder": "Jos ei valittu, tiedostotulostetta ei luoda.",
-    "embeddings-tab-file-output-directory-textbox-label": "Tiedostotulosteen hakemisto",
-    "embeddings-select-file-output-directory-button-label": "Valitse tiedostotulosteen hakemisto",
     "eval-tab-accuracy-checkbox-info": "Tarkkuus mittaa mallin oikeiden ennusteiden prosenttiosuutta.",
     "eval-tab-annotation-col-accordion-label": "Merkintäsarakkeet",
     "eval-tab-annotation-selection-button-label": "Valitse merkintäkansio",
@@ -240,6 +240,7 @@
     "species-list-coordinates-week-slider-info": "Määritä viikko, jolloin äänitys tehtiin, käyttäen yksinkertaistettua järjestelmää, jossa jokainen kuukausi on jaettu neljään viikkoon. Valitse arvo väliltä 1-48.",
     "species-list-coordinates-week-slider-label": "Viikko",
     "species-list-coordinates-yearlong-checkbox-label": "Ympärivuotinen",
+    "species-list-custom-classifier-no-labelfile-warning": "Label-tiedostoa muodossa <customclassifier>_Labels.txt ei löytynyt.",
     "species-list-custom-classifier-selection-button-label": "Valitse luokittelija",
     "species-list-custom-list-file-label": "Tiedosto",
     "species-list-radio-info": "Suodata lajit, jotka näytetään tuloksissa.",

birdnet_analyzer/lang/fr.json

@@ -2,6 +2,7 @@
     "analyze-locale-dropdown-info": "Locale pour les noms communs des espèces traduits dans les résultats",
     "analyze-locale-dropdown-label": "Locale",
     "analyze-start-button-label": "Analyser",
+    "embedding-file-output-accordion-label": "Paramètres de sortie de fichier",
     "embedding-settings-accordion-label": "Paramètres",
     "embedding-settings-audio-speed-slider-info": "Les valeurs négatives ralentissent, les valeurs positives accélèrent. La modification de la vitesse audio sera stockée dans la base de données.",
     "embedding-settings-audio-speed-slider-label": "Modification de la vitesse audio",
@@ -41,18 +42,17 @@
     "embeddings-search-select-query-button-label": "Sélectionner l'échantillon de requête",
     "embeddings-search-start-button-label": "Démarrer la recherche",
     "embeddings-search-tab-title": "Recherche",
+    "embeddings-select-file-output-directory-button-label": "Sélectionner le répertoire de sortie de fichier",
     "embeddings-tab-dataset-info": "Spécifiez le nom du jeu de données",
     "embeddings-tab-dataset-label": "Jeu de données",
     "embeddings-tab-db-info": "Le nom de la base de données",
+    "embeddings-tab-file-output-directory-textbox-label": "Répertoire de sortie de fichier",
+    "embeddings-tab-file-output-directory-textbox-placeholder": "Si non sélectionné, la sortie de fichier sera omise.",
     "embeddings-tab-finish-info": "Extraction des incorporations terminée vers :",
     "embeddings-tab-select-db-directory-button-label": "Sélectionner le répertoire de la base de données",
     "embeddings-tab-select-input-directory-button-label": "Sélectionner le répertoire d'entrée (récursif)",
     "embeddings-tab-start-button-label": "Extraire les incorporations",
     "embeddings-tab-title": "Incorporations",
-    "embedding-file-output-accordion-label": "Paramètres de sortie de fichier",
-    "embeddings-tab-file-output-directory-textbox-placeholder": "Si non sélectionné, la sortie de fichier sera omise.",
-    "embeddings-tab-file-output-directory-textbox-label": "Répertoire de sortie de fichier",
-    "embeddings-select-file-output-directory-button-label": "Sélectionner le répertoire de sortie de fichier",
     "eval-tab-accuracy-checkbox-info": "La précision mesure le pourcentage de prédictions correctes faites par le modèle.",
     "eval-tab-annotation-col-accordion-label": "Colonnes d'annotation",
     "eval-tab-annotation-selection-button-label": "Sélectionner le répertoire d'annotations",
@@ -240,6 +240,7 @@
     "species-list-coordinates-week-slider-info": "Spécifiez la semaine de l'année où l'enregistrement a été effectué, en utilisant un système simplifié où chaque mois est divisé en quatre semaines. Choisissez une valeur de 1 à 48.",
     "species-list-coordinates-week-slider-label": "Semaine",
     "species-list-coordinates-yearlong-checkbox-label": "Toute l'année",
+    "species-list-custom-classifier-no-labelfile-warning": "Aucun fichier de labels trouvé au format <customclassifier>_Labels.txt.",
     "species-list-custom-classifier-selection-button-label": "Sélectionner un classificateur",
     "species-list-custom-list-file-label": "Fichier",
     "species-list-radio-info": "Filtrer les espèces qui sont incluses dans le résultat.",