Merged commit includes the following changes: (#7249)

* Merged commit includes the following changes:
257930561  by yongzhe:

    Mobile LSTD TfLite Client.

--
257928126  by yongzhe:

    Mobile SSD Tflite client.

--
257921181  by menglong:

    Fix discrepancy between pre_bottleneck = {true, false}

--
257561213  by yongzhe:

    File utils.

--
257449226  by yongzhe:

    Mobile SSD Client.

--
257264654  by yongzhe:

    SSD utils.

--
257235648  by yongzhe:

    Proto bazel build rules.

--
256437262  by Menglong Zhu:

    Fix check for FusedBatchNorm op to only verify it as a prefix.

--
256283755  by yongzhe:

    Bazel build and copybara changes.

--
251947295  by yinxiao:

    Add missing interleaved option in checkpoint restore.

--
251513479  by yongzhe:

    Conversion utils.

--
248783193  by yongzhe:

    Branch protos needed for the lstd client.

--
248200507  by menglong:

    Fix proto namespace in example config

--

P...

research/lstm_object_detection/tflite/mobile_lstd_tflite_client.cc

@@ -98,9 +98,13 @@ bool MobileLSTDTfLiteClient::InitializeInterpreter(
     return false;
   }
 
-  // Outputs are: raw_outputs/box_encodings, raw_outputs/class_predictions,
+  // Outputs are:
+  //   TFLite_Detection_PostProcess,
+  //   TFLite_Detection_PostProcess:1,
+  //   TFLite_Detection_PostProcess:2,
+  //   TFLite_Detection_PostProcess:3,
   //   raw_outputs/lstm_c, raw_outputs/lstm_h
-  if (interpreter_->outputs().size() != 4) {
+  if (interpreter_->outputs().size() != 6) {
     LOG(ERROR) << "Invalid number of interpreter outputs: " <<
         interpreter_->outputs().size();
     return false;
@@ -108,12 +112,12 @@ bool MobileLSTDTfLiteClient::InitializeInterpreter(
 
   const std::vector<int> output_tensor_indices = interpreter_->outputs();
   const TfLiteTensor& output_lstm_c =
-      *interpreter_->tensor(output_tensor_indices[2]);
+      *interpreter_->tensor(output_tensor_indices[4]);
   if (!ValidateStateTensor(output_lstm_c, "output lstm_c")) {
     return false;
   }
   const TfLiteTensor& output_lstm_h =
-      *interpreter_->tensor(output_tensor_indices[3]);
+      *interpreter_->tensor(output_tensor_indices[5]);
   if (!ValidateStateTensor(output_lstm_h, "output lstm_h")) {
     return false;
   }
@@ -121,6 +125,8 @@ bool MobileLSTDTfLiteClient::InitializeInterpreter(
   // Initialize state with all zeroes.
   lstm_c_data_.resize(lstm_state_size_);
   lstm_h_data_.resize(lstm_state_size_);
+  lstm_c_data_uint8_.resize(lstm_state_size_);
+  lstm_h_data_uint8_.resize(lstm_state_size_);
 
   if (interpreter_->AllocateTensors() != kTfLiteOk) {
     LOG(ERROR) << "Failed to allocate tensors";
@@ -201,6 +207,7 @@ bool MobileLSTDTfLiteClient::QuantizedInference(const uint8_t* input_data) {
   CHECK(input_data) << "Input data cannot be null.";
   uint8_t* input = interpreter_->typed_input_tensor<uint8_t>(0);
   CHECK(input) << "Input tensor cannot be null.";
+  memcpy(input, input_data, input_size_);
 
   // Copy input LSTM state into TFLite's input tensors.
   uint8_t* lstm_c_input = interpreter_->typed_input_tensor<uint8_t>(1);
@@ -215,14 +222,18 @@ bool MobileLSTDTfLiteClient::QuantizedInference(const uint8_t* input_data) {
   CHECK_EQ(interpreter_->Invoke(), kTfLiteOk) << "Invoking interpreter failed.";
 
   // Copy LSTM state out of TFLite's output tensors.
-  // Outputs are: raw_outputs/box_encodings, raw_outputs/class_predictions,
+  // Outputs are:
+  //   TFLite_Detection_PostProcess,
+  //   TFLite_Detection_PostProcess:1,
+  //   TFLite_Detection_PostProcess:2,
+  //   TFLite_Detection_PostProcess:3,
   //   raw_outputs/lstm_c, raw_outputs/lstm_h
-  uint8_t* lstm_c_output = interpreter_->typed_output_tensor<uint8_t>(2);
+  uint8_t* lstm_c_output = interpreter_->typed_output_tensor<uint8_t>(4);
   CHECK(lstm_c_output) << "Output lstm_c tensor cannot be null.";
   std::copy(lstm_c_output, lstm_c_output + lstm_state_size_,
             lstm_c_data_uint8_.begin());
 
-  uint8_t* lstm_h_output = interpreter_->typed_output_tensor<uint8_t>(3);
+  uint8_t* lstm_h_output = interpreter_->typed_output_tensor<uint8_t>(5);
   CHECK(lstm_h_output) << "Output lstm_h tensor cannot be null.";
   std::copy(lstm_h_output, lstm_h_output + lstm_state_size_,
             lstm_h_data_uint8_.begin());

research/lstm_object_detection/tflite/mobile_ssd_client.cc

@@ -76,7 +76,7 @@ bool MobileSSDClient::BatchDetect(
       LOG(ERROR) << "Post Processing not supported.";
       return false;
     } else {
-      if (NoPostProcessNoAnchors(detections[batch])) {
+      if (!NoPostProcessNoAnchors(detections[batch])) {
         LOG(ERROR) << "NoPostProcessNoAnchors failed.";
         return false;
       }