Input nodes missing in frozen graph

With TensorFlow v1.9.0, I went through the training instructions and produced the following .pbtxt and checkpoint files:

celoss.JPG
model.ckpt-104063.meta
model.ckpt-120000.data-00000-of-00002
checkpoint
model.ckpt-111901.data-00000-of-00002  
model.ckpt-120000.data-00001-of-00002
events.out.tfevents.1558022250.indy    
model.ckpt-111901.data-00001-of-00002  
model.ckpt-120000.index
graph.pbtxt                            
model.ckpt-111901.index                
model.ckpt-120000.meta
locloss.JPG                           
model.ckpt-111901.meta                 
model.ckpt-96213.data-00000-of-00002
loss.JPG                               
model.ckpt-119735.data-00000-of-00002  
model.ckpt-96213.data-00001-of-00002
model.ckpt-104063.data-00000-of-00002  
model.ckpt-119735.data-00001-of-00002 
model.ckpt-96213.index
model.ckpt-104063.data-00001-of-00002  
model.ckpt-119735.index                
model.ckpt-96213.meta
model.ckpt-104063.index                
model.ckpt-119735.meta                 
predict

I then tried to freeze the graph:

$ freeze_graph \
    --input_graph=./logs/graph.pbtxt \
    --input_binary=false \
    --input_checkpoint=./logs/model.ckpt-120000 \
    --output_graph=frozen_graph.pb \
    --output_node_names=ssd300/Reshape_13,ssd300/Reshape_12

2019-05-22 13:25:28.443832: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1392] Found device 0 with properties: 
name: GeForce GTX 1080 major: 6 minor: 1 memoryClockRate(GHz): 1.8475
pciBusID: 0000:07:00.0
totalMemory: 7.92GiB freeMemory: 7.35GiB
2019-05-22 13:25:28.443862: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1471] Adding visible gpu devices: 0
2019-05-22 13:25:28.766253: I tensorflow/core/common_runtime/gpu/gpu_device.cc:952] Device interconnect StreamExecutor with strength 1 edge matrix:
2019-05-22 13:25:28.766284: I tensorflow/core/common_runtime/gpu/gpu_device.cc:958]      0 
2019-05-22 13:25:28.766292: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0:   N 
2019-05-22 13:25:28.766383: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1084] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 7091 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1080, pci bus id: 0000:07:00.0, compute capability: 6.1)

This seemed to successfully produce a frozen graph file: 105203994 May 21 13:55 frozen_graph.pb

When I inspected the frozen_graph.pb file, it appears that input nodes are missing from the graph:

$ decent_q inspect --input_frozen_graph frozen_graph.pb 

Op types used: 167 Const, 71 Identity, 35 BiasAdd, 35 Conv2D, 23 Relu, 15 Reshape, 12 Transpose, 12 StridedSlice, 12 Pack, 5 MaxPool, 2 Mul, 2 ConcatV2, 1 Maximum, 1 QueueDequeueManyV2, 1 FIFOQueueV2, 1 Rsqrt, 1 SpaceToBatchND, 1 Square, 1 Sum, 1 BatchToSpaceND

No inputs found.
Found 2 possible outputs: (name=ssd300/Reshape_12, op=Reshape) (name=ssd300/Reshape_13, op=Reshape)

Any ideas as to why the input node(s) are missing? What should the input nodes for this model be?

Thank you.

HiKapok / SSD.TensorFlow

Input nodes missing in frozen graph #114