diff --git a/networks/ssd_inception_v2_gesture/Makefile b/networks/ssd_inception_v2_gesture/Makefile
index fc013df4..a1a5c26e 100644
--- a/networks/ssd_inception_v2_gesture/Makefile
+++ b/networks/ssd_inception_v2_gesture/Makefile
@@ -5,7 +5,7 @@ YELLOW='\033[1;33m'
 NOCOLOR='\033[0m'
 RED = '\033[1;31m'
 
-GRAPH_FILENAME_BASE = ssd_inception_v2_food
+GRAPH_FILENAME_BASE = ssd_inception_v2_gesture
 GRAPH_DIR_FP32 = ssd_inception_v2_mo_fp32
 GRAPH_DIR_FP16 = ssd_inception_v2_mo_fp16
 GRAPH_FILENAME_BASE_IN_DIR_FP32 = ${GRAPH_DIR_FP32}/frozen_inference_graph.xml
@@ -28,10 +28,44 @@ MODEL_INFERENCE_PB = frozen_inference_graph.pb
 GET_MODEL_PIPELINE = wget -c --no-cache -P ./tensorflow_model https://raw.githubusercontent.com/fcr3/gesture_detection/master/tensorflow_model/${MODEL_INFERENCE_PIPELINE_CONFIG}
 GET_MO_MODEL_FP32_LABELS = wget -c --no-cache -P ./ssd_inception_v2_mo_fp32 https://raw.githubusercontent.com/fcr3/gesture_detection/master/model_optimized_fp32/${MO_LABELS}
 GET_MO_MODEL_FP16_LABELS = wget -c --no-cache -P ./ssd_inception_v2_mo_fp16 https://raw.githubusercontent.com/fcr3/gesture_detection/master/model_optimized_fp16/${MO_LABELS}
-GET_MODEL_PB = python3 download_weights.py
+GET_MODEL_PB = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/model/frozen_inference_graph.pb -O ${MODEL_DIR}/${MODEL_INFERENCE_PB}
+# if above doesn't work: python3 download_weights.py
 
 RUN_PY_RELATIVE_DIR = ssd_inception_v2_gesture.py
 
+TRAIN_DIR = train_files
+TRAIN_CHKPT_DIR = ${TRAIN_DIR}/ssd_inception_v2_coco_2018_01_28
+PRE_TRAINED_MODEL = wget -c --no-cache -P ./${TRAIN_DIR} http://download.tensorflow.org/models/object_detection/ssd_inception_v2_coco_2018_01_28.tar.gz
+
+TRAIN_RECORD_FILE = ./${TRAIN_DIR}/train.record
+TRAIN_RECORD = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/training/train.record -O ${TRAIN_RECORD_FILE}
+# TRAIN_RECORD_LINK = 'https://drive.google.com/uc?id=13uM9aoxrUvPjZTIkXv33av_Dx8CYhg7S'
+
+TEST_RECORD_FILE = ./${TRAIN_DIR}/test.record
+TEST_RECORD = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/training/test.record -O ${TEST_RECORD_FILE}
+
+TRAIN_PBTXT_FILE = ./${TRAIN_DIR}/object-detection.pbtxt
+TRAIN_PBTXT = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/training/object-detection.pbtxt -O ${TRAIN_PBTXT_FILE}
+
+SSD_CONFIG_FILE = ./${TRAIN_DIR}/ssd_inception_v2_coco.config
+SSD_CONFIG = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/training/ssd_inception_v2_coco.config -O ${SSD_CONFIG_FILE}
+
+TRAIN_PIPE_CONFIG_FILE = ./${TRAIN_DIR}/pipeline.config
+TRAIN_PIPE_CONFIG = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/training/pipeline.config -O ${TRAIN_PIPE_CONFIG_FILE}
+
+JPEG_ZIP_FILE = ./${TRAIN_DIR}/JPEGImages.zip
+JPEG_DOWNLOAD = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/training/JPEGImages-20200116T001612Z-001.zip -O ${JPEG_ZIP_FILE}
+# JPEG_FOLDER_LINK = 'https://drive.google.com/uc?id=1J5p63WjnkcWAyCQVbaW6qshsUBfX7aZ2'
+
+ANNO_ZIP_FILE = ./${TRAIN_DIR}/Annotations.zip
+ANNO_DOWNLOAD = wget http://dp2f0iayzbweh.cloudfront.net/ncappzoo/gesture_detection/training/Annotations-20200115T233047Z-001.zip -O ${ANNO_ZIP_FILE}
+# ANNO_FOLDER_LINK = 'https://drive.google.com/uc?id=17ZmWWQFV5ljMWTX1HP-dZtD25xKXVCYU'
+
+NETWORK_NAME = SSD Inception v2: Gesture Model
+TF_MODEL_DIRNAME = models/
+TF_MODEL_REPO = git clone https://github.com/tensorflow/models.git
+PROTOC_FILE = models/research/protobuf.zip
+PROTOC_DOWNLOAD = wget -O ${PROTOC_FILE} https://github.com/google/protobuf/releases/download/v3.0.0/protoc-3.0.0-linux-x86_64.zip
 
 .PHONY: all
 all: deps data compile_model
@@ -78,10 +112,280 @@ get_model:
 		fi ; \
 	fi
 
+.PHONY: compile_from_training
+compile_from_training:
+	@echo $(YELLOW) '\n'${NETWORK_NAME}": Information disclosure before compiling model." $(NOCOLOR);
+	@echo "The following steps will be taken in order to compile your own model that detects produce:"
+	@echo "  - Freeze trained model"
+	@echo "  - Run frozen model through model optimizer"
+	@echo ""
+	@echo $(YELLOW)"Before proceeding, edit the compile_from_training script in the Makefile."
+	@echo "Specifically, edit the trained_checkpoin_prefix to reflect your true (latest) model checkpoint."$(NOCOLOR)
+	@read -p"Press ENTER to continue the steps or Ctrl-C to cancel" variable_nps; echo;
+	@if [ -e ${TF_MODEL_DIRNAME}research/object_detection/export_gesture_graph ] ;\
+	then \
+		echo "Exported graph exists. Skipping exporting graph..." ;\
+	else \
+		echo "Models directory for TF Object Detection was not found" ;\
+		echo "Please downloading training files using: make download_train_files" ;\
+    cd ${TF_MODEL_DIRNAME}research ;\
+    ./bin/protoc object_detection/protos/*.proto --python_out=. ;\
+    export PYTHONPATH=$PYTHONPATH:`pwd`/object_detection ;\
+    export PYTHONPATH=$PYTHONPATH:`pwd`:`pwd`/slim ;\
+    cd object_detection ;\
+    mkdir export_gesture_graph ;\
+    python3 export_inference_graph.py \
+    --input_type=image_tensor \
+    --pipeline_config_path=training_gesture/ssd_inception_v2_coco.config \
+    --trained_checkpoint_prefix=training_gesture/model.ckpt-0 \
+    --output_directory=export_gesture_graph ;\
+	fi
+	@echo "Compiling model using model optimizer..."
+	@if [ -e ${TF_MODEL_DIRNAME}research/object_detection/export_gesture_graph ] ;\
+	then \
+    if [ -z "$(INTEL_OPENVINO_DIR)" ] ; \
+    then \
+            echo "Please initiate the OpenVINO environment before optimizing TF graph."; \
+    else \
+            echo "OpenVINO environment set. Continuing with optimizing TF graph..." ;\
+						cd ${TF_MODEL_DIRNAME}research/object_detection/export_gesture_graph ;\
+						mkdir ssd_inception_v2_mo_fp32_train ;\
+						mkidr ssd_inception_v2_mo_fp16_train ;\
+						${NCCOMPILE} \
+						--data_type=FP32 --reverse_input_channels \
+						--input_model frozen_inference_graph.pb \
+						--tensorflow_object_detection_api_pipeline_config ../training_gesture/ssd_inception_v2_coco.config \
+						--tensorflow_use_custom_operations_config ../../../../ssd_v2_support_inception_train.json \
+						--output_dir ssd_inception_v2_mo_fp32_train ;\
+						${NCCOMPILE} \
+						--data_type=FP16 --reverse_input_channels \
+						--input_model frozen_inference_graph.pb \
+						--tensorflow_object_detection_api_pipeline_config ../training_gesture/ssd_inception_v2_coco.config \
+						--tensorflow_use_custom_operations_config ../../../../ssd_v2_support_inception_train.json \
+						--output_dir ssd_inception_v2_mo_fp16_train ;\
+						mv ssd_inception_v2_mo_fp16_train ../../../../ssd_inception_v2_mo_fp16_train ;\
+						mv ssd_inception_v2_mo_fp32_train ../../../../ssd_inception_v2_mo_fp32_train ;\
+						echo "Successfully exported and optimized training graphs!" ;\
+    fi ; \
+	else \
+		echo "No exported graph was made, so graph optimization could not be completed." ;\
+	fi
 
 .PHONY: train
-train:
-	@echo $(YELLOW) "\nSSD Inception V2 Geseture: Training model... NOT YET IMPLEMENTED" $(NOCOLOR);
+train: train_info download_train_files set_up_training_api copy_training_files train_model
+
+.PHONY: train_info
+train_info:
+	@echo $(YELLOW) '\n'${NETWORK_NAME}": Information disclosure before training model." $(NOCOLOR);
+	@echo "The following steps will be taken in order to train your own model that detects produce:"
+	@echo "  - Download training files"
+	@echo "  - Set up and locally install training environment and API"
+	@echo "  - Copy training files into desired location for training"
+	@echo "  - Train model using Tensorflow Object Detection API"
+	@echo ""
+	@echo "Throughout this whole process, files will be downloaded and repositories will be cloned."
+	@echo $(YELLOW)"Please pay attention to the prompts before proceeding as installation may error"
+	@echo "if not followed properly."$(NOCOLOR)
+	@read -p"Press ENTER to continue the steps above (this might take a long time) or Ctrl-C to cancel" variable_nps; echo;
+
+.PHONY: download_train_files
+download_train_files:
+	@echo $(YELLOW)"Downloading train files..."
+	@if [ -e ${TRAIN_CHKPT_DIR} ] ; \
+	then \
+		echo $(NOCOLOR)"Checkpoint files already exist, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing checkpoint files from Tensorflow..."; \
+		${PRE_TRAINED_MODEL} ;\
+		tar -xvzf ${TRAIN_CHKPT_DIR}.tar.gz ;\
+		mv ssd_inception_v2_coco_2018_01_28 ${TRAIN_CHKPT_DIR} ;\
+	fi
+	@if [ -e ${TEST_RECORD_FILE} ] ; \
+	then \
+		echo $(NOCOLOR)"Test record file already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing test record file..."; \
+		${TEST_RECORD} ;\
+	fi
+	@if [ -e ${TRAIN_PBTXT_FILE} ] ; \
+	then \
+		echo $(NOCOLOR)"Pbtxt file already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing pbtxt file..."; \
+		${TRAIN_PBTXT} ;\
+	fi
+	@if [ -e ${SSD_CONFIG_FILE} ] ; \
+	then \
+		echo $(NOCOLOR)"SSD Inception v2 config already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing SSD Inception v2 config file..."; \
+		${SSD_CONFIG} ;\
+	fi
+	@if [ -e ${TF_MODEL_DIRNAME} ] ; \
+	then \
+		echo $(NOCOLOR)"Models repo already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Cloning models repo..."; \
+		${TF_MODEL_REPO} ;\
+	fi
+	@if [ -e ${PROTOC_FILE} ] ; \
+	then \
+		echo $(NOCOLOR)"Protoc file already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing protoc file..."; \
+		${PROTOC_DOWNLOAD} ;\
+	fi
+	@echo "Downloading larger files using gdown, checking for installation"
+	@if [ -e ${ANNO_ZIP_FILE} ] ; \
+	then \
+		echo $(NOCOLOR)"Annotated zip file already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing annotated zip file..."; \
+		${ANNO_DOWNLOAD} ; \ # gdown ${ANNO_FOLDER_LINK} -O ${ANNO_ZIP_FILE} ;\ ;\
+		unzip ${ANNO_ZIP_FILE} -d train_files/ ;\
+	fi
+	@if [ -e ${TRAIN_RECORD_FILE} ] ; \
+	then \
+		echo $(NOCOLOR)"Train record already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing train record file..."; \
+		${TRAIN_RECORD} ; \ # gdown ${TRAIN_RECORD_LINK} -O train_files/train.recor ;\
+	fi
+	@if [ -e ${JPEG_ZIP_FILE} ] ; \
+	then \
+		echo $(NOCOLOR)"JPEG zip file already exists, skipping network request."; \
+	else \
+		echo $(NOCOLOR)"Grabbing JPEG zip file..."; \
+		${JPEG_DOWNLOAD} ; \ # gdown ${JPEG_FOLDER_LINK} -O train_files/JPEGImages.zi ;\
+		unzip ${JPEG_ZIP_FILE} -d train_files/ ;\
+	fi
+	@if [ -e cocoapi/PythonAPI ] ;\
+	then \
+		echo $(NOCOLOR)"cocoapi repo already exists, skipping network request."; \
+	else \
+		echo "Cloning cocoapi repo for future installation..." ;\
+		git clone https://github.com/cocodataset/cocoapi.git ;\
+	fi
+	@echo ""
+
+.PHONY: set_up_training_api
+set_up_training_api:
+	@echo $(YELLOW)"Setting up training environment..."
+	@if [ -e cocoapi/PythonAPI ] ; \
+	then \
+		echo $(YELLOW)"URGENT: READ FOLLOWING BEFORE PROCEEDING:" ;\
+		echo "Replacing 'python' with 'python3' in the Makefile within cocoapi/PythonAPI"$(NOCOLOR) ;\
+		read -p"Press ENTER to proceed or CTRL+C to exit..." variable_nps; echo ;\
+		echo $(NOCOLOR)"Installing COCO API..." ;\
+		cd cocoapi/PythonAPI ;\
+		sed -i 's/python/python3/g' Makefile ;\
+		make ;\
+		cp -r pycocotools ../../models/research/ ;\
+	else \
+		echo $(NOCOLOR)"Please downloading training files before set up."; \
+	fi
+	@if [ -e ${TF_MODEL_DIRNAME} ] ; \
+	then \
+		echo $(NOCOLOR)"Models repo exists, proceeding with setup."; \
+		echo $(NOCOLOR)"Setting up protobuf."; \
+		cd models/research/ ;\
+		unzip protobuf.zip ;\
+		./bin/protoc object_detection/protos/*.proto --python_out=. ;\
+		export PYTHONPATH=$PYTHONPATH:`pwd`/object_detection ;\
+		export PYTHONPATH=$PYTHONPATH:`pwd`:`pwd`/slim ;\
+	else \
+		echo $(NOCOLOR)"Please downloading training files before set up."; \
+	fi
+	@echo $(YELLOW)"Testing training environment..."
+	@if python3 -c 'import pkgutil; exit(not pkgutil.find_loader("tensorflow"))' ; \
+	then \
+		echo $(NOCOLOR)'Tensorflow found, proceeding with testing the training environment...' ; \
+		cd models/research/ ;\
+		./bin/protoc object_detection/protos/*.proto --python_out=. ;\
+		export PYTHONPATH=$PYTHONPATH:`pwd`/object_detection ;\
+		export PYTHONPATH=$PYTHONPATH:`pwd`:`pwd`/slim ;\
+		python3 object_detection/builders/model_builder_test.py ;\
+	else \
+		echo 'Tensorflow not found. Please install tensorflow (CPU version).' ; \
+		echo 'If you would like to install Tensorflow-gpu, please analyze your' ; \
+		echo 'your systems hardware and install the required drivers. Then, pip/pip3' ; \
+		echo 'install tensorflow-gpu (might need special permissions).' ; \
+	fi
+	@echo ""
+
+.PHONY: copy_training_files
+copy_training_files:
+	@echo $(YELLOW)"Moving training files..."
+	@echo $(NOCOLOR)"Will potentially overwrite training files in models/research/object_detection"
+	@echo "Please check if there are training files within models/research/object_detection."
+	@read -p"Press ENTER to proceed or CTRL+C to exit..." variable_nps; echo;
+	@echo $(NOCOLOR)"Checking if training files directory exists in current directory..."
+	@if [ -e ${TRAIN_DIR} ] ;\
+	then \
+		echo $(NOCOLOR)"Training files exist, proceeding to copy them into object detection directory."; \
+		cp -a ${TRAIN_DIR}/ssd_inception_v2_coco_2018_01_28 models/research/object_detection/. ; \
+		mkdir models/research/object_detection/training_gesture ;\
+		cp -a ${TRAIN_DIR}/. models/research/object_detection/training_gesture/. ;\
+		echo "Moved training files into models/research/object_detection" ;\
+	else \
+		echo $(NOCOLOR)"Please download training files."; \
+	fi
+	@echo ""
+
+
+.PHONY: train_model
+train_model:
+	@echo $(YELLOW) '\n'${NETWORK_NAME}": Training model..." $(NOCOLOR);
+	@echo "  - Trained network  will be saved in:";
+	@echo "      training_gesture";
+	@echo ""
+	@echo $(YELLOW)"URGENT: READ THE FOLLOWING INSTRUCTIONS BEFORE TRAINING:" ;
+	@echo "In order to have more verbose logging so you can verify that your model is" ;
+	@echo "actually training, adding the following to line 57 of model_main.py in" ;
+	@echo "models/research/object_detection: tf.logging.set_verbosity(tf.logging.INFO)."
+	@echo "A reference file will be made with the original file content: old_model_main.py" $(NOCOLOR);
+	@echo ""
+	@echo "NOTE: Training files are unfrozen. If would like to export a model,";
+	@echo "enter 'make export_and_optimize_model' to generate an optimized model";
+	@echo "from your Tensorflow model.";
+	@echo ""
+	@echo "NOTE: If your system is running out of memory, please decrease batch size to 1.";
+	@echo "Work upwards to empirically test if your system can handle increased batch sizes.";
+	@echo "To edit batch size, edit the ssd_inception_v2_coco.config file in models/research/training_gesture by"
+	@echo "replacing the number on line 136 to any positive integer (ex. 1) after 'batch_size:'."
+	@echo ""
+	@read -p"Press ENTER to continue the steps above (this might take a long time) or Ctrl-C to cancel" variable_nps; echo;
+
+	@if [ -e ${TF_MODEL_DIRNAME}/research/object_detection/old_model_main.py ] ;\
+	then \
+		echo "Changing model_main.py" ;\
+		cd models/research/object_detection ;\
+		rm -rf model_main.py ;\
+		cp old_model_main.py model_main.py ;\
+		sed -i -e '57itf.logging.set_verbosity(tf.logging.INFO)' model_main.py ;\
+	else \
+		echo "Changing model_main.py and adding a reference file" ;\
+		cd models/research/object_detection ;\
+		cp model_main.py old_model_main.py ;\
+		sed -i -e '57itf.logging.set_verbosity(tf.logging.INFO)' model_main.py ;\
+	fi
+	@if [ -e ${TF_MODEL_DIRNAME} ] ; \
+	then \
+		echo "Directory: ${TF_MODEL_DIRNAME}research/object_detection exists, setting up environment..."; \
+		cd models/research/ ;\
+		./bin/protoc object_detection/protos/*.proto --python_out=. ;\
+		export PYTHONPATH=$PYTHONPATH:`pwd`/object_detection ;\
+		export PYTHONPATH=$PYTHONPATH:`pwd`:`pwd`/slim ;\
+		cd object_detection ;\
+		echo "Environment is setup, commencing with training..." ;\
+		python3 model_main.py \
+    --pipeline_config_path=training_gesture/ssd_inception_v2_coco.config \
+    --model_dir=training_gesture/ \
+    --num_train_steps=20000 \
+    --alsologtostderr ;\
+	else \
+		echo "Cannot locate the training files for TF Object Detection. Please download."; \
+	fi ;
 
 
 .PHONY: compile_model
@@ -118,15 +422,20 @@ compile_model: get_model
 		${GET_MO_MODEL_FP16_LABELS} ;\
 	fi
 
+.PHONY: install-reqs
+install-reqs:
+	@echo $(YELLOW)"\nSSD Inception V2 Gesture: Checking application requirements...\n"$(NOCOLOR)
+	@echo "Nothing to install..."
+
 
 .PHONY: run_FP16
-run_FP16: deps data compile_model
+run_FP16: install-reqs deps data compile_model
 	@echo $(YELLOW) "\nSSD Inception V2: Running Python sample..." $(NOCOLOR)
 	python3 $(RUN_PY_RELATIVE_DIR) -i 'cam' -m ${GRAPH_FILENAME_BASE_IN_DIR_FP16} --labels ${MO_LABELS_IN_DIR_FP16} -d MYRIAD
 
 
 .PHONY: run_FP32
-run_FP32: deps data compile_model
+run_FP32: install-reqs deps data compile_model
 	@echo $(YELLOW) "\nSSD Inception V2: Running Python sample..." $(NOCOLOR)
 	python3 $(RUN_PY_RELATIVE_DIR) -i 'cam' -m ${GRAPH_FILENAME_BASE_IN_DIR_FP32} -l ${CPU_EXTENSION} --labels ${MO_LABELS_IN_DIR_FP32}
 
@@ -136,19 +445,13 @@ run: run_py
 
 
 .PHONY: run_py
-run_py: deps data compile_model run_FP16
+run_py: install-reqs deps data compile_model run_FP16
 
 
-.PHONY: install-reqs
-install-reqs:
-	@echo $(YELLOW)"\nSSD Inception V2 Gesture: Checking application requirements...\n"$(NOCOLOR)
-	@echo "No requirements needed."
-	
-	
 .PHONY: uninstall-reqs
-uninstall-reqs: 
+uninstall-reqs:
 	@echo $(YELLOW)"\nSSD Inception V2 Gesture: Uninstalling requirements..."$(NOCOLOR)
-	@echo "Nothing to uninstall."
+	@echo "\nNo requirements to install..."
 
 
 .PHONY: help
@@ -173,3 +476,8 @@ clean: clean
 	rm -rf ${GRAPH_DIR_FP32}
 	rm -rf ${GRAPH_DIR_FP16}
 	rm -rf ${MODEL_DIR}
+	rm -rf train_files
+	rm -rf models
+	rm -rf cocoapi
+	rm -rf ssd_inception_v2_mo_fp16_train
+	rm -rf ssd_inception_v2_mo_fp32_train
diff --git a/networks/ssd_inception_v2_gesture/README.md b/networks/ssd_inception_v2_gesture/README.md
index 65c089e1..3ce4f9e9 100644
--- a/networks/ssd_inception_v2_gesture/README.md
+++ b/networks/ssd_inception_v2_gesture/README.md
@@ -45,7 +45,7 @@ Runs the `ssd_inception_v2_gesture.py` script with the FP32 network. Note that t
 Runs the `ssd_inception_v2_gesture.py` script with the FP16 network. Users must plug in their Intel Neural Compute Stick 2 in order to successfully run this application.
 
 ### make train
-**TO BE IMPLEMENTED.** Trains a SSD Inception V2 model using the Tensorflow Object Detection API given an `Annotations` and `JPEGImages` folder containing .xml and .jpg images, respectively, for training. Training is not necessary since the sample will download a pre-trained model. This option allows for the user to further refine the SSD Inception V2 model if they so desire.
+Trains a SSD Inception V2 model using the Tensorflow Object Detection API given an `Annotations` and `JPEGImages` folder containing .xml and .jpg images, respectively, for training. Training is not necessary since the sample will download a pre-trained model. This option allows for the user to further refine the SSD Inception V2 model if they so desire.
 
 ### make help
 Shows makefile possible targets and brief descriptions.
diff --git a/networks/ssd_inception_v2_gesture/download_weights.py b/networks/ssd_inception_v2_gesture/download_weights.py
deleted file mode 100644
index 63550004..00000000
--- a/networks/ssd_inception_v2_gesture/download_weights.py
+++ /dev/null
@@ -1,23 +0,0 @@
-import requests
-
-def main():
-    pb_file_id = '1TuK2SMtkFlT5SRrhjWCz4Z5xXE-gcRDd'
-    file_to_write = './tensorflow_model/frozen_inference_graph.pb'
-    drive_url = "https://docs.google.com/uc?export=download"
-
-    req_session = requests.Session()
-    session_res = req_session.get(drive_url, params = { 'id' : pb_file_id }, stream = True)
-
-    for k, v in session_res.cookies.items():
-        if k.startswith('download_warning'):
-            params = { 'id' : pb_file_id, 'confirm' : v }
-            session_res = session_res.get(drive_url, params = params, stream = True)
-            break
-
-    size = 32768
-    with open(file_to_write, "wb") as f:
-        for packet in session_res.iter_content(size):
-            f.write(packet) if packet else None
-
-if __name__ == "__main__":
-    main()
diff --git a/networks/ssd_inception_v2_gesture/ssd_v2_support_inception_train.json b/networks/ssd_inception_v2_gesture/ssd_v2_support_inception_train.json
new file mode 100644
index 00000000..f9892bcf
--- /dev/null
+++ b/networks/ssd_inception_v2_gesture/ssd_v2_support_inception_train.json
@@ -0,0 +1,62 @@
+[
+    {
+        "custom_attributes": {
+        },
+        "id": "ObjectDetectionAPIPreprocessorReplacement",
+        "inputs": [
+            [
+                {
+                    "node": "map/Shape$",
+                    "port": 0
+                },
+                {
+                    "node": "map/TensorArrayUnstack/Shape$",
+                    "port": 0
+                },
+                {
+                    "node": "map/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3$",
+                    "port": 2
+                }
+            ]
+        ],
+        "instances": [
+            ".*Preprocessor/"
+        ],
+        "match_kind": "scope",
+        "outputs": [
+            {
+                "node": "sub$",
+                "port": 0
+            },
+            {
+                "node": "map/TensorArrayStack_1/TensorArrayGatherV3$",
+                "port": 0
+            }
+        ]
+    },
+    {
+        "custom_attributes": {
+            "code_type": "caffe.PriorBoxParameter.CENTER_SIZE",
+            "pad_mode": "caffe.ResizeParameter.CONSTANT",
+            "resize_mode": "caffe.ResizeParameter.WARP"
+        },
+        "id": "ObjectDetectionAPISSDPostprocessorReplacement",
+        "include_inputs_to_sub_graph": true,
+        "include_outputs_to_sub_graph": true,
+        "instances": {
+            "end_points": [
+                "detection_boxes",
+                "detection_scores",
+                "num_detections"
+            ],
+            "start_points": [
+                "Postprocessor/Shape",
+                "Postprocessor/scale_logits",
+                "Postprocessor/Tile",
+                "Postprocessor/Reshape_1",
+                "Postprocessor/Cast_1"
+            ]
+        },
+        "match_kind": "points"
+    }
+]