Fixes issue #3, #5, #6 and #8

README.md

@@ -77,6 +77,8 @@ Both model is trained using the [Tensorflow Object Detection](https://github.com
 
 ## Anime related repos and datasets
 
+[Danbooru2019 Figures](https://www.gwern.net/Crops#danbooru2019-figures) is a dataset generated by [Gwern](https://github.com/gwern) who ran AniSeg on Danbooru2019 solo SFW images to generate 855k crops of close-up single-character whole-body/profile images. Check out a BIGGAN model trained on that dataset [here](https://www.gwern.net/Faces#danbooru2019e621-256px-biggan).
+
 Shameless self promotion of my [TwinGAN](https://github.com/jerryli27/TwinGAN) model to turn people into anime characteres and cats!
 
 Sketch coloring using [PaintsTransfer](https://github.com/lllyasviel/style2paints) and [PaintsChainer](http://paintschainer.preferred.tech/).

infer_from_image.py

@@ -20,6 +20,12 @@
                        'Path to the output TFRecord.')
 tf.flags.DEFINE_boolean('visualize_inference', False,
                         'If set, also outputs the annotated inference result image.')
+tf.flags.DEFINE_boolean('output_cropped_image', False,
+                        'If set, also outputs the cropped image to the output path. e.g. '
+                        'OUTPUT_PATH/IMAGE_NAME_crop.png.')
+tf.flags.DEFINE_boolean('only_output_cropped_single_object', False,
+                        'Only used if FLAGS.output_cropped_image is True. Only outputs the cropped image if there is '
+                        'one and only one object detected.')
 
 tf.flags.DEFINE_string('inference_graph', None,
                        'Path to the inference graph with embedded weights.')
@@ -89,6 +95,25 @@ def main(_):
       min_score_thresh=FLAGS.min_score_thresh,
       visualize_inference=FLAGS.visualize_inference,
       feed_dict={image_ph: image_np})
+
+    if FLAGS.output_cropped_image:
+      if FLAGS.only_output_cropped_single_person and len(result["detection_score"]) == 1:
+        num_outputs = 1
+      else:
+        num_outputs = len(result["detection_score"])
+
+      for crop_i in range(0, num_outputs):
+        if (result["detection_score"])[crop_i] > FLAGS.min_score_thresh:
+          base, ext = os.path.splitext(os.path.basename(image_path))
+          output_crop = os.path.join(FLAGS.output_path, base + '_crop_%d.png' %crop_i)
+          idims = image_np.shape  # np array with shape (height, width, num_color(1, 3, or 4))
+          min_x = int(min(round(result["detection_bbox_xmin"][crop_i] * idims[1]), idims[1]))
+          max_x = int(min(round(result["detection_bbox_xmax"][crop_i] * idims[1]), idims[1]))
+          min_y = int(min(round(result["detection_bbox_ymin"][crop_i] * idims[0]), idims[0]))
+          max_y = int(min(round(result["detection_bbox_ymax"][crop_i] * idims[0]), idims[0]))
+          image_cropped = image_np[min_y:max_y, min_x:max_x, :]
+          util_io.imsave(output_crop, image_cropped)
+
     if FLAGS.visualize_inference:
       output_image = os.path.join(FLAGS.output_path, os.path.basename(image_path))
       util_io.imsave(output_image, result['annotated_image'])

object_detection/inference/detection_inference.py

@@ -97,7 +97,7 @@ def build_inference_graph(image_tensor, inference_graph_path, override_num_detec
     detected_labels_tensor: Detected labels. Int64 tensor,
         shape=[num_detections]
   """
-  with tf.gfile.Open(inference_graph_path, 'r') as graph_def_file:
+  with tf.gfile.Open(inference_graph_path, 'rb') as graph_def_file:
     graph_content = graph_def_file.read()
   graph_def = tf.GraphDef()
   graph_def.MergeFromString(graph_content)

object_detection/inference/mask_inference.py

@@ -117,7 +117,7 @@ def build_inference_graph(image_tensor, inference_graph_path, override_num_detec
     detected_labels_tensor: Detected labels. Int64 tensor,
         shape=[num_detections]
   """
-  with tf.gfile.Open(inference_graph_path, 'r') as graph_def_file:
+  with tf.gfile.Open(inference_graph_path, 'rb') as graph_def_file:
     graph_content = graph_def_file.read()
   graph_def = tf.GraphDef()
   graph_def.MergeFromString(graph_content)

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_015_v79HxnWIuNY-0075.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.15097324550151825], "detection_class_label": [1], "detection_bbox_xmin": [0.02185318060219288], "detection_bbox_xmax": [0.33384037017822266], "detection_bbox_ymax": [1.0], "detection_score": [0.9998868703842163]}
+{"detection_bbox_ymin": [0.15097332000732422], "detection_class_label": [1], "detection_bbox_xmin": [0.021853169426321983], "detection_bbox_xmax": [0.33384037017822266], "detection_bbox_ymax": [1.0], "detection_score": [0.9998868703842163]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_017_7eo6Cq1Wcmw-0054.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0, 0.0, 0.0], "detection_class_label": [1, 1, 1], "detection_bbox_xmin": [0.13196967542171478, 0.17053768038749695, 0.4854376018047333], "detection_bbox_xmax": [0.9733457565307617, 0.6400503516197205, 0.9591410756111145], "detection_bbox_ymax": [1.0, 1.0, 1.0], "detection_score": [0.9990824460983276, 0.9750493764877319, 0.8864908814430237]}
+{"detection_bbox_ymin": [0.0, 0.0, 0.0], "detection_class_label": [1, 1, 1], "detection_bbox_xmin": [0.1319696307182312, 0.17053766548633575, 0.48543766140937805], "detection_bbox_xmax": [0.9733458757400513, 0.6400503516197205, 0.9591410756111145], "detection_bbox_ymax": [1.0, 1.0, 1.0], "detection_score": [0.9990824460983276, 0.9750492572784424, 0.8864911794662476]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_019_-FC7r2xpHeM-0020.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.031232278794050217, 0.02029275894165039, 0.08043289184570312], "detection_class_label": [1, 1, 1], "detection_bbox_xmin": [0.057558923959732056, 0.25248056650161743, 0.04003767669200897], "detection_bbox_xmax": [0.6568614840507507, 0.612417995929718, 0.38541293144226074], "detection_bbox_ymax": [0.9163234233856201, 0.9481703042984009, 0.7146922945976257], "detection_score": [0.9985986351966858, 0.9692012071609497, 0.8951168060302734]}
+{"detection_bbox_ymin": [0.031232357025146484, 0.02029275894165039, 0.08043287694454193], "detection_class_label": [1, 1, 1], "detection_bbox_xmin": [0.057558901607990265, 0.25248056650161743, 0.0400376096367836], "detection_bbox_xmax": [0.656861424446106, 0.612417995929718, 0.38541293144226074], "detection_bbox_ymax": [0.9163235425949097, 0.9481701850891113, 0.7146923542022705], "detection_score": [0.9985986351966858, 0.9692013263702393, 0.8951167464256287]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_020_XNf_K-86qfc-0015.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0, 0.0], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.009168218821287155, 0.36362728476524353], "detection_bbox_xmax": [0.5205694437026978, 0.9845380187034607], "detection_bbox_ymax": [1.0, 1.0], "detection_score": [0.8680392503738403, 0.8005604147911072]}
+{"detection_bbox_ymin": [0.0, 0.0], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.009168196469545364, 0.36362722516059875], "detection_bbox_xmax": [0.5205694437026978, 0.9845380187034607], "detection_bbox_ymax": [1.0, 1.0], "detection_score": [0.868039071559906, 0.800560474395752]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_020_XNf_K-86qfc-0053.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.06623085588216782], "detection_class_label": [1], "detection_bbox_xmin": [0.35986754298210144], "detection_bbox_xmax": [0.7026284337043762], "detection_bbox_ymax": [1.0], "detection_score": [0.9999532699584961]}
+{"detection_bbox_ymin": [0.06623077392578125], "detection_class_label": [1], "detection_bbox_xmin": [0.35986754298210144], "detection_bbox_xmax": [0.7026284337043762], "detection_bbox_ymax": [1.0], "detection_score": [0.9999532699584961]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_023_NAvuPXQlsHc-0038.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.6799054145812988], "detection_class_label": [1], "detection_bbox_xmin": [0.8207776546478271], "detection_bbox_xmax": [0.9607893228530884], "detection_bbox_ymax": [0.9834651947021484], "detection_score": [0.9982274174690247]}
+{"detection_bbox_ymin": [0.679905354976654], "detection_class_label": [1], "detection_bbox_xmin": [0.8207776546478271], "detection_bbox_xmax": [0.9607893228530884], "detection_bbox_ymax": [0.983465313911438], "detection_score": [0.9982274174690247]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_024_BkknXGYLZjg-0027.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.4395271837711334], "detection_bbox_xmax": [0.8358895778656006], "detection_bbox_ymax": [0.9876407980918884], "detection_score": [0.9329289793968201]}
+{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.4395271837711334], "detection_bbox_xmax": [0.8358895778656006], "detection_bbox_ymax": [0.9876407980918884], "detection_score": [0.9329290390014648]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_032_T66-dT9j12E-0028.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.008851885795593262], "detection_class_label": [1], "detection_bbox_xmin": [0.22670131921768188], "detection_bbox_xmax": [0.7149162888526917], "detection_bbox_ymax": [0.8868505358695984], "detection_score": [0.9998735189437866]}
+{"detection_bbox_ymin": [0.008851846680045128], "detection_class_label": [1], "detection_bbox_xmin": [0.22670136392116547], "detection_bbox_xmax": [0.7149162292480469], "detection_bbox_ymax": [0.8868504762649536], "detection_score": [0.9998735189437866]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_033_6ic7vtLu27o-0077.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.056579314172267914], "detection_class_label": [1], "detection_bbox_xmin": [0.28377458453178406], "detection_bbox_xmax": [0.623889148235321], "detection_bbox_ymax": [0.9893053770065308], "detection_score": [0.9998255372047424]}
+{"detection_bbox_ymin": [0.056579314172267914], "detection_class_label": [1], "detection_bbox_xmin": [0.28377461433410645], "detection_bbox_xmax": [0.623889148235321], "detection_bbox_ymax": [0.9893053770065308], "detection_score": [0.9998255372047424]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_040_up3W6vwnqWE-0017.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0, 0.05746563524007797], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.0, 0.6614444851875305], "detection_bbox_xmax": [0.36700305342674255, 0.9948204159736633], "detection_bbox_ymax": [1.0, 0.9930106997489929], "detection_score": [0.9999117851257324, 0.9991965889930725]}
+{"detection_bbox_ymin": [0.0, 0.05746559426188469], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.0, 0.6614445447921753], "detection_bbox_xmax": [0.36700305342674255, 0.9948202967643738], "detection_bbox_ymax": [1.0, 0.9930106997489929], "detection_score": [0.9999117851257324, 0.9991965889930725]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_040_up3W6vwnqWE-0058.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.052236080169677734], "detection_class_label": [1], "detection_bbox_xmin": [0.37460678815841675], "detection_bbox_xmax": [0.6751424074172974], "detection_bbox_ymax": [1.0], "detection_score": [0.999936580657959]}
+{"detection_bbox_ymin": [0.052235763520002365], "detection_class_label": [1], "detection_bbox_xmin": [0.37460678815841675], "detection_bbox_xmax": [0.6751424074172974], "detection_bbox_ymax": [1.0], "detection_score": [0.999936580657959]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_042_KI2r0WTbatw-0016.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.09422492980957031, 0.10073117911815643, 0.05561808869242668], "detection_class_label": [1, 1, 1], "detection_bbox_xmin": [0.12375533580780029, 0.5590329170227051, 0.05892935395240784], "detection_bbox_xmax": [0.39927706122398376, 0.9086330533027649, 0.49791550636291504], "detection_bbox_ymax": [0.9686238765716553, 1.0, 1.0], "detection_score": [0.9999614953994751, 0.9998766183853149, 0.8046053647994995]}
+{"detection_bbox_ymin": [0.09422492980957031, 0.10073117911815643, 0.05561816692352295], "detection_class_label": [1, 1, 1], "detection_bbox_xmin": [0.12375533580780029, 0.5590329766273499, 0.058929331600666046], "detection_bbox_xmax": [0.39927706122398376, 0.9086330533027649, 0.49791550636291504], "detection_bbox_ymax": [0.9686238765716553, 1.0, 1.0], "detection_score": [0.9999614953994751, 0.9998766183853149, 0.8046048283576965]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_048_NdEswAUqE-8-0020.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.27610674500465393], "detection_bbox_xmax": [0.7200565338134766], "detection_bbox_ymax": [1.0], "detection_score": [0.5710054039955139]}
+{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.27610674500465393], "detection_bbox_xmax": [0.7200565338134766], "detection_bbox_ymax": [1.0], "detection_score": [0.5710049867630005]}

samples/figure_detection_outputs/PL0bHKk6wuUGL_Qd34mf0XsQnyiDk2OeGR_059_Tajzm4ai73M-0040.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.2724817991256714], "detection_bbox_xmax": [0.7115412354469299], "detection_bbox_ymax": [1.0], "detection_score": [0.9999473094940186]}
+{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.272481769323349], "detection_bbox_xmax": [0.7115411758422852], "detection_bbox_ymax": [1.0], "detection_score": [0.9999473094940186]}

samples/figure_detection_outputs/s - 799902.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0, 0.031649600714445114], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.059222374111413956, 0.03007768653333187], "detection_bbox_xmax": [0.8767802119255066, 0.6009855270385742], "detection_bbox_ymax": [0.9823748469352722, 0.9407722353935242], "detection_score": [0.9999626874923706, 0.9133853912353516]}
+{"detection_bbox_ymin": [0.0, 0.031649671494960785], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.05922233313322067, 0.030077628791332245], "detection_bbox_xmax": [0.8767802715301514, 0.600985586643219], "detection_bbox_ymax": [0.9823748469352722, 0.9407721161842346], "detection_score": [0.9999626874923706, 0.9133856892585754]}

samples/figure_detection_outputs/s - 799903.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.2590962052345276, 0.4136689603328705, 0.10286876559257507, 0.5568650960922241, 0.06504856050014496, 0.3132608234882355, 0.3796249032020569, 0.2886377274990082], "detection_class_label": [1, 1, 1, 1, 1, 1, 1, 1], "detection_bbox_xmin": [0.0, 0.5071926116943359, 0.7001814246177673, 0.3440193831920624, 0.7081868052482605, 0.030533110722899437, 0.42722249031066895, 0.0008578062406741083], "detection_bbox_xmax": [0.32470256090164185, 0.5619336366653442, 0.9357579350471497, 0.40758606791496277, 0.948326051235199, 0.3081423342227936, 0.5113341808319092, 0.19651828706264496], "detection_bbox_ymax": [0.9045862555503845, 0.6943526864051819, 0.5952692627906799, 0.7578580975532532, 0.8525521755218506, 0.7585433721542358, 0.530319094657898, 0.9236617684364319], "detection_score": [0.9995057582855225, 0.9983882904052734, 0.9963197708129883, 0.9932727813720703, 0.9767559766769409, 0.9588579535484314, 0.6218403577804565, 0.541231095790863]}
+{"detection_bbox_ymin": [0.25909629464149475, 0.4136689603328705, 0.10286876559257507, 0.5568650960922241, 0.06504852324724197, 0.3132607638835907, 0.3796249032020569, 0.2886377274990082], "detection_class_label": [1, 1, 1, 1, 1, 1, 1, 1], "detection_bbox_xmin": [0.0, 0.5071926116943359, 0.7001814842224121, 0.3440193831920624, 0.7081868052482605, 0.03053312376141548, 0.42722249031066895, 0.000857794308103621], "detection_bbox_xmax": [0.32470250129699707, 0.5619336366653442, 0.9357579350471497, 0.40758606791496277, 0.948326051235199, 0.30814236402511597, 0.5113341808319092, 0.19651827216148376], "detection_bbox_ymax": [0.904586136341095, 0.6943526864051819, 0.5952692627906799, 0.7578580975532532, 0.8525521755218506, 0.7585433721542358, 0.530319094657898, 0.9236617684364319], "detection_score": [0.9995057582855225, 0.9983882904052734, 0.9963197708129883, 0.9932727813720703, 0.9767559766769409, 0.9588578343391418, 0.6218402981758118, 0.5412302017211914]}

samples/figure_detection_outputs/s - 799904.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.0], "detection_bbox_xmax": [1.0], "detection_bbox_ymax": [0.9407608509063721], "detection_score": [0.7478146553039551]}
+{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.0], "detection_bbox_xmax": [1.0], "detection_bbox_ymax": [0.9407609105110168], "detection_score": [0.7478148341178894]}

samples/figure_detection_outputs/s - 799906.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.07171551883220673], "detection_class_label": [1], "detection_bbox_xmin": [0.21395057439804077], "detection_bbox_xmax": [0.9700915813446045], "detection_bbox_ymax": [0.8001722693443298], "detection_score": [0.9970467686653137]}
+{"detection_bbox_ymin": [0.07171560078859329], "detection_class_label": [1], "detection_bbox_xmin": [0.21395064890384674], "detection_bbox_xmax": [0.9700915813446045], "detection_bbox_ymax": [0.8001722693443298], "detection_score": [0.9970467686653137]}

samples/figure_detection_outputs/s - 799907.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.016685273498296738, 0.10914835333824158], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.013866310007870197, 0.020841311663389206], "detection_bbox_xmax": [0.982010006904602, 0.6000389456748962], "detection_bbox_ymax": [0.9424865245819092, 0.9282892346382141], "detection_score": [0.9985717535018921, 0.8433741927146912]}
+{"detection_bbox_ymin": [0.016685212031006813, 0.10914816707372665], "detection_class_label": [1, 1], "detection_bbox_xmin": [0.013866386376321316, 0.020841330289840698], "detection_bbox_xmax": [0.982010006904602, 0.6000389456748962], "detection_bbox_ymax": [0.9424865245819092, 0.9282891154289246], "detection_score": [0.9985717535018921, 0.8433746099472046]}

samples/figure_detection_outputs/s - 799910.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.07372306287288666], "detection_bbox_xmax": [0.9705819487571716], "detection_bbox_ymax": [0.914916455745697], "detection_score": [0.9999948740005493]}
+{"detection_bbox_ymin": [0.0], "detection_class_label": [1], "detection_bbox_xmin": [0.07372302561998367], "detection_bbox_xmax": [0.9705819487571716], "detection_bbox_ymax": [0.914916455745697], "detection_score": [0.9999948740005493]}

samples/figure_detection_outputs/s - 799911.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.020250430330634117, 0.21967953443527222, 0.13471992313861847, 0.7732296586036682, 0.8389378786087036, 0.060042064636945724], "detection_class_label": [1, 1, 1, 1, 1, 1], "detection_bbox_xmin": [0.6443371176719666, 0.12162468582391739, 0.3648078739643097, 0.11651556938886642, 0.35263320803642273, 0.372955322265625], "detection_bbox_xmax": [0.7593987584114075, 0.2412860095500946, 0.5030063986778259, 0.23377394676208496, 0.4853338599205017, 0.5026745200157166], "detection_bbox_ymax": [0.27595147490501404, 0.4020392596721649, 0.2688125967979431, 1.0, 1.0, 0.2917357385158539], "detection_score": [0.964841365814209, 0.9447920918464661, 0.9354586005210876, 0.6816180348396301, 0.6629226207733154, 0.5766783356666565]}
+{"detection_bbox_ymin": [0.020250430330634117, 0.21967953443527222, 0.13471993803977966, 0.773229718208313, 0.8389378786087036, 0.060042090713977814], "detection_class_label": [1, 1, 1, 1, 1, 1], "detection_bbox_xmin": [0.6443371176719666, 0.12162467837333679, 0.3648078739643097, 0.11651555448770523, 0.35263320803642273, 0.372955322265625], "detection_bbox_xmax": [0.7593987584114075, 0.2412860095500946, 0.5030063986778259, 0.23377397656440735, 0.4853339195251465, 0.5026745200157166], "detection_bbox_ymax": [0.27595147490501404, 0.4020392596721649, 0.2688126266002655, 1.0, 1.0, 0.2917357087135315], "detection_score": [0.964841365814209, 0.9447922110557556, 0.9354584813117981, 0.6816179156303406, 0.662921667098999, 0.5766789317131042]}

samples/figure_detection_outputs/s - 799913.json

@@ -1 +1 @@
-{"detection_bbox_ymin": [0.18048936128616333], "detection_class_label": [1], "detection_bbox_xmin": [0.38603949546813965], "detection_bbox_xmax": [0.7949660420417786], "detection_bbox_ymax": [0.7883065342903137], "detection_score": [0.999916672706604]}
+{"detection_bbox_ymin": [0.18048924207687378], "detection_class_label": [1], "detection_bbox_xmin": [0.38603949546813965], "detection_bbox_xmax": [0.7949660420417786], "detection_bbox_ymax": [0.7883065342903137], "detection_score": [0.999916672706604]}